Between chat discussion and the various edits to the question, I have come to the conclusion that, more than anything else, your suggestions represent premature optimization—you are trying to preemptively fix problems that haven’t happened yet and you are unfamiliar with.
Pathfinder is a system with a lot of design problems, particularly as one scales up in level. But design problems only represent risks of actual problems. There are tables that play Pathfinder up to high levels and never notice any problems—either through sheer dumb luck, or from simply being sufficiently wrapped up in the enjoyment of the story, characters, and social activity of the game (read: things that no system can really claim credit for, but are ultimately the reason we play) to be bothered by the problems. There are many other tables for whom the design problems pose real, serious problems at the table, at which point the design problems of the system are inhibiting the table’s enjoyment of the game—but even then, plenty of tables will enjoy themselves despite the system, because the good things about the game overshadow those issues.
The reason to be aware of design issues, then, is to make your life easier—to be aware of the pitfalls ahead of you, and to find the source of problems you do encounter. Sometimes the answer to “why is my Pathfinder game having this problem?” is “because of some flaw in its design, which you might improve by...” And then, having experienced that problem or ones like it, understanding the system, you can change the design to try to eliminate some of these flaws.
But in your case, you are playing your first Pathfinder campaign. It is just starting out, and you are still at mid-low levels. The problems with the system, so far, have not been problems at your table. The risk of these design problems becoming problems at the table will increase as you go up in level, and at the highest levels (to say nothing of the post-20th levels you propose), the risks become large enough that most tables will have some problems, but right now, you’re doing fine.
That means you are trying to fix problems you think you might have—and I think you might have them too, but ultimately it’s impossible to judge an appropriate fix for a problem you haven’t yet seen. The best advice anyone here can give you, then, is I think that you should play your game, wait and see how things turn out, and adjust things as necessary when and if problems actually occur. It’s well-and-good to avoid problems in the first place, but in this case you’re too far out ahead of the problems and so cannot judge the appropriate remedy.
Nonetheless, since you asked, here are my thoughts on your suggestions, and how they do or do not address the design problems of Pathfinder. By addressing a design problem, I mean how well they limit or eliminate the risk of running into actual problems at the table, ideally while at the same time making more character styles viable and reducing the difficulty of building a functional character and avoiding character-creation traps.
However, having recently read https://rpg.stackexchange.com/a/9054/33272 (fighter player said casters have a harder time at high levels), I am not sure I am going down the right road any more.
The post you link from that fighter is flat-out wrong. It’s just not true. I’ve downvoted it, but there’s nothing more I can do. The fact remains, it’s a bad answer, no matter how well-written or how much people wish it were true, because it’s simply inaccurate. Sorry.
The only true way to fix the caster-martial imbalance is to go through all the spells, and rewrite most of them. Good luck with that.
- Classes that have access to 9th level spells eventually (Druid/Cleric/Sorcerer/Wizard/etc) would use the spell progression tables from Dnd 5e.
The 5e spell progression change is a good idea, but it’s not enough. The very existence of some of those high-level spells is in-and-of-itself a problem (and personally, I am not inclined to do enough work myself to determine exactly which).
- I would use the limited magic system
Limited magic is an awful rule. It will certainly hurt spellcasters a lot, but it also causes a lot of problems with the game’s math (you will basically never use any spell that allows a save ever, because odds are your target will save). Plenty of spells also only become relevant with higher-than-minimum caster level (like the ever-popular grease), which now just become traps.
Which would all be fine (spellcasters can still certainly be extremely powerful even in this system), but I don’t like the warping it causes. All of a sudden, all kinds of iconic and popular narrative tropes for spellcasters become impossibilities. Spellcasters are forced to all use the few spells that remain amazing even under those rules—which means you still have overpowered spellcasters, you just have less variety of them. That’s not great. Even if you can get around the limits of limited magic with Heighten Spell (which still works poorly because the game’s math assumes ability scores much higher than the minimums to cast a spell), why should anyone when they could just use other spells?
And that is the real problem, and the reason why no broad-stroke band-aid fix is going to solve the problem—there will always be another spell that is still overpowered under your new rules. Instead, these kinds of approaches just cause wonky problems, skew what internal balance does exist, and fail to actually fix the original problems completely.
The rule smells to me very strongly of something written by someone who heard spellcasters were overpowered, but didn’t actually understand the problems. Someone who didn’t want to do the hard work of going through every spell, and wanted an easy way out.
There is no easy way out. There is no way to apply a simple, broad-stroke rule to Pathfinder to fix its balance problems. There is no way these problems will ever be fully handled short of actually going through every spell in the game.
- I would allow the use of Esoteric Components
Personally, I consider estoteric components just the epitome of un-fun, and would probably avoid using them. I haven’t gone through the list of components, but I would also be leery that there are some that are overpowered, just adding to problems with spellcasters.
- I have doubled the cost of meta-magic rods, and made them full-round actions to use.
OK; it’s not a terrible idea but it’s not going to fix the big picture.
- I have just disallowed summoners all together.
The unchained summoner is actually a nerf, and is at least better than the original one is. Ultimately, as unreasonable as the (original) summoner was, though, cleric and wizard and so on were always still much better.
- As a side note, I have made the changes suggested here: Balancing Spellcasters; Spontaneous Casters versus Prepared Casters, in which spontaneous casters can use an immediate action to use metamagic, and use the same progression as wizards.
A note: that was written before the arcanist was. The arcanist is an absurdly overpowered class with the only caveat being the delayed spellcasting progresion. An arcanist without that suddenly becomes clearly and without question the strongest class in the game.
Honestly I suggest you just ban it, too.
To be honest, I am surprised how well the numbers seem to be working out! The bottom line is a lot of killer spells are getting moved into the 6th-9th level slot area, to still be killer, and this is where slots have been reduced the most by using the 5e slots.
Those aren’t “the killer spells.” In fact, of those you list, black tentacles is the only one that’s even good, much less killer.
And no, I am not going to counter it with my own list of killer spells. That isn’t the point, and would only produce more arguments.
So my question is: Is there really a problem I need to be fixing via house-rules (with the aim of class balance and keeping the game playable), and if so would the above rule changes fix it, or at least make it more balanced and playable?
Pathfinder as a system breaks down on its own and without any particular intent on doing so somewhere around 15th level. As in, it becomes difficult to keep the game workable even when people are trying to. There really isn’t a lot that can be done about it without a serious overhaul to the system: houserules, any houserules, are not really sufficient to the task.
What you propose here is a series of broad-stroke band-aid fixes, some good, some bad. Overall, is it an improvement? Probably. Does it actually fix the problems? Emphatically no, it does not.
The best way to avoid these problems in Pathfinder, in fact the only way to avoid these problems in Pathfinder that I have found, is to simply not allow it to happen in the first place: sticking to low levels tends to result in a much better game. Pathfinder is still imbalanced even at the lowest levels, but the problem is less than it otherwise would be.
If you are really intent on a big, reality-changing epic campaign, you can still do it at relatively low levels. For example, in D&D 3.5e, the Eberron campaign setting mostly maxes out (barring the continent of deus ex machina dragons) at around 12th or 13th level. That means you only have to be 10th-ish to be operating at the highest echelons of power in that setting.
Alternatively, the E6 houserule is quite popular. The idea is that because 7th-level introduces 4th-level spells (which are the beginning of the “mundanes can’t even be in the same room” levels of power disparity in Pathfinder), we stop at 6th level. However, characters continue to gain XP, and every basically gain bonus feats at certain XP thresholds (when they otherwise would have leveled up, or every \$x\$ XP, or whatever).
I have heard of games using similar concepts, but stopping later than 6th. I would suggest that 12th is about as high as I would recommend, but that can work too. I have also recently started an interesting game wherein people start as a 6th-level commoner, and then further levels are added as with gestalt characters, so the levels you take “upgrade” the commoner levels without actually resulting in higher level (we haven’t gotten that far yet, but the plan is that when we would have otherwise hit 7th level, we would actually add yet another layer of gestalt).
Finally, Pathfinder as a system works much better if you restrict the classes available to a certain “power range.” This is the point of the tier system, which applies at least as much to Pathfinder. In fact, the best games of Pathfinder I’ve played are ones in which everyone was playing a class that receives 6th-level spells. This is a decent “sweet spot” where everyone gets “enough” magic to be competitive in the extremely-high-magic world of Pathfinder, and no one gets “too much” magic where they are bringing in far-too-powerful spells.
But no, I don’t think that Pathfinder works well as a system past approximately 15th level. From my own anecdotal experience, the games in which it does are either being played by people who know the system very well, and are carefully not breaking it, or groups in which everyone plays at 15th level as if the game hadn’t changed since 5th level—that is to say, people are ignoring the huge growth in power available and going with old standbyes that are long-since obsolete. But the former is difficult to do, and the latter is a delicate, fragile balance easily destroyed by someone trying something new, and also seems to ask why you are even bothering with high-level characters in the first place.