“But wait, evil clerics can’t get spontaneous cure!”
This answer has drawn numerous comments claiming that a good cleric’s spontaneous cure is important to survival, evil clerics must now prepare cure spells since they lack that feature, etc. etc. This is an understandable sentiment: that is how the game depicts clerics, casting cure spells round after round. Certainly, if you’re going to be using cure a lot, casting it spontaneously is the way to go, since that allows you to prepare any other answers you want. So the ability to spontaneously cast cure spells is supposed to be a large advantage; it’s described as such and it’s very likely that its authors (going all the way back to Wizards of the Coast’s original implementation) intended it to be such.
But it’s just not true. This is the “trap” of the cleric class.
I call it a “trap” because the description of the cleric class leads you to thinking it’s the right way to play the class, and because once you start down that path, you can get stuck there.
So what’s the problem here? The cure spells are bad. Their numbers are just too small, which means you need to combine them with channeling positive energy, you need to get magic items and take feats, all to try and get them into a good place. And all of those options that you can get are scaled based on the cure spells, so you spend a lot of gold, or a nearly-priceless feat, on small boosts to what you can do. And you spend all your time trying to keep up with cure spells that are never going to cut it.
(Note that almost all of this analysis is specific to cure spells—the heal and mass heal spells do not follow the same trends, because those spells have very good numbers, not to mention powerful effects beyond hp healing, and so are legitimately strong choices. But those spells are also unaffected by alignment, since the good cleric’s ability to spontaneously convert prepared spells into cure spells doesn’t let them do the same for heal or mass heal. You just have to prepare those, but they’re easily strong enough to be worth preparing. They are also strong enough not to need extra feats or magic items to boost them.)
In combat, the most valuable thing you have are actions. You need to do as much as possible to turn the tide in your party’s favor with every single one, because you will not get many. Characters can do a lot on their turns, and so can the creatures they fight. A typical Pathfinder combat is decided in 2-3 turns, tops—there might be mopping up to do, or a retreat to manage, but whether or not you are ultimately going to win is usually pretty clear after that point. That should make you suspicious of the relatively low numbers on cure spells—and you’d be right to be so.
It is very difficult to cure someone of more damage than a typical foe of the same level can do in a single attack. Plenty of characters and creatures can have multiple attacks right from 1st level, and just about everybody who cares about attacking has them by 8th at the latest. So when you spend your turn casting cure, your turn has been spent undoing less than an enemy turn’s worth of damage. That’s a losing proposition.
Moreover, the cleric is a formidable fighter—which means they can easily deal more damage than they can cure, particularly if they build for it. And the cleric spell list is very, very good—which gives you lots of other options for protecting your party or dealing with enemies.
So you could inefficiently cure, and spend an entire turn undoing a portion of an enemy’s turn, or you can be more proactive, and do something that directly eliminates or mitigates threats, either by limiting enemies’ ability to attack or by enhancing allies’ defenses (or indirectly, by buffing allies’ offense so they can limit enemies’ ability to attack by killing things). In pure numbers, the latter approach saves vastly more hp than the former can heal in the same amount of time. Finally, remember that healing is itself inherently inefficient—you need to wait until the enemy has dealt damage in order to heal it. That means you always risk the enemy killing someone in between—not good.
All of which means, when your few opportunities to determine the outcome of a fight come up, it is going to be very rare that a cure spell is the best possible answer. So why do people swear by it? Because they’ve built clerics who are focused on cure spells, and now they struggle to do much else. The cleric is extremely versatile, but it is possible to get yourself into a hole with one—and focusing on cure spells is a good way to do that, because cure requires so many resources focused on it to keep up. It’s entirely possible that these clerics can’t do better than cure, and that without the contribution of a fully-powered cleric to protect people or power through the fights faster, in-combat healing becomes necessary. But that’s a losing proposition every time it happens, and it is far stronger to build a cleric around the idea that it’s not going to—that we’re going to get through this fight, and then heal. At which point the wand of cure light wounds or wand of infernal healing is the way to go.
This actually kind of reflects a “thing” with Pathfinder, and D&D 3.5e before it: the things that the authors thought were good, often aren’t. Because they thought they were good, they were conservative with them. Oftentimes, too conservative—as with cure spells. Most direct-damage spells are in a similar boat, for the record (though their numbers aren’t nearly as bad as cure or inflict spells’). So very often the best options aren’t the ones that are presented as the best options, because those options were reined in and other options—not perceived as strong—were buffed to make them seem more appealing. This kind of over-correction is found throughout both systems.