Do you really need the constant time pressure?
But slowing down, or stopping in a city to "prepare" for more than a few days makes no sense.
Is this statement an accurate reflection of just how acute the time pressure is in your game?
I'm not saying I don't believe you. I do. But stop for a second and figure out whether time pressure is a critical part of the structure of the campaign or a dial you can adjust easily as the GM.
Oftentimes, as modern people living in a world of awesome infrastructure, we forget how long stuff takes. Getting food, going places, acquiring goods. Historical armies, for example, move overland very slowly — and it could take months to gather your banners and arrange the necessary provisions and support for any kind of significant offensive move.
Even in a highly-magical context, where real-life constraints like how long it takes to march from one city to another are irrelevant, bad guys can have plans that take years to come to fruition. They may include prominent stages where the PCs just have to wait as well — to see what their enemy's next move, to wait for some event that both the heroes and villains are going to need to try to exploit for their benefit. Removing logistical realism doesn't mean the game has to plunge ahead at a breakneck pace; it just means time can move at "the speed of plot."
Do the system-imposed time constraints actually matter?
Serious limitations on what you can do with your time can be a key part of the system's overall resource model. It's important to understand these before you modify them.
In your particular case, D&D 3rd Edition does try to use time as a resource constraint, but mostly on the scale of combat turns or "per day" resources; the system really doesn't care whether days or weeks or months pass between the individual adventures.
Also, I would argue that D&D 3rd Edition really doesn't take "realistic" time constraints particularly seriously. This is a game where any injury, up to and including death, only sets characters back about a day, and where travel becomes instantaneous at higher levels!
The crafting times for skill use are an entirely different part of D&D3, designed to approximate economic activity that's largely irrelevant to the adventures the characters are having in play. If the PCs want to use their skills to sit around and make money, then bring up the crafting times. But if they're trying to use their normally-time-consuming skills to help themselves out of a tight spot during their doomsday-racing magical adventures, then accept that they're doing extraordinary things and throw out time requirements designed for "downtime" color rather than drama.
Replace time-based restrictions with a different kind of obstacle!
Sometimes, time-based restrictions or complications can be part of the challenge. "Can we accomplish this before something else happens?" is good for dramatic tension, especially since it's an easy way to create challenges that "fail forward" — transforming into a new situation rather than a frustrating roadblock if the protagonists don't actually succeed at their task.
Time-based restrictions on most tasks seem to be a poor fit for your particular campaign. Because they amount to just outright saying "No." In the context of the game you're playing, "Should I skip the next three adventures so that I can learn my new spells?" is not an interesting or meaningful choice. So, take those kinds of tasks and turn them into a different kind of challenge.
Honestly, if it isn't an interesting challenge (either in the narrative sense or the strategic sense) at all, just give it to them. If you do want the activity to represent overcoming a real impediment, make it one that can be overcome within the context of this particular story.
Example applications
- An in-the-moment adventure obstacle: For activities that represent overcoming the obstacles in front of you, like making a crude raft to help you cross the river or using your research skill to find a map to the next dungeon, most systems have you as the GM are setting the timetable for tests anyway. So just pick a duration that doesn't mess up the flow of the narrative. If it means that a task that normally takes 2d4 days takes 2 hours, who cares?
- Abilities with built-in time-wasters: For stuff that's supposed to just force a bit of downtime, like wizards learning spells, you can have it happen automatically between each adventure. If you don't want to waive the limitations completely, you can modify the rules to count days but not require characters to spend all their time devoted to the task. Maybe it still takes two days to learn that new spell, for example, but you can count "active time" against that — they still have to manage their time but it doesn't force a hard stop on their other activities.
- Consumables: If PCs are using crafting to refill their basic consumables, it's reasonable to allow them the equivalent of a single test between each adventure "for free," just kinda folded into the basic recovery downtime in the game. If the outcome of the test is interesting, you can make a small skill challenge out of it: if they succeed, they get to restock their resources; if they fail, they're going in with depleted supplies. The price they have to pay is that they can't use this process to stockpile stuff — if they already have a full supply, they don't get to make surplus stuff without rolling a more "normal" (time-sensitive) check. This is essentially crafting "at the speed of plot."
- Significant long-term effects: An action that has lasting effects on the game, like crafting a powerful magic item, can involve a "side quest" to procure rare and special magical materials instead of the normal lengthy downtime. This still takes up some time but it's "active" time and everybody can participate.