Adding in anything published by WotC for settings adds in an astonishing 240-odd for a total of just under 600 prestige classes. (Again, before counting Dragon content, which was official...)
OK, round it up to a thousand, then. The 3.5 PH sold something north of 300k units, so there were at least that many players. Does it really seem so likely that at least /one/ of those 300k might've played any given one of those 1k PrCs?
Maybe not played it very long or enjoyed it very much, but at least wrote it on a character sheet and sat down to play at least once before regretting it. ;P
Balance in 5e is more fluid than 4e, but is still much tighter than 3e. Option creep equally broke the power curve of both prior editions.
Not exactly. 5e's balance is more fluid than either of the other modern editions, because it is so DM-Empowering, and puts class balance squarely in the DM's hands. The DM makes what he wants of it. That's what I was trying to get across briefly as 'fluid.' 3.5, according to Mr. Cook, anyway, was designed with the specific intent of building in 'rewards for system mastery' that were, in essence, intentionally imbalanced. It wasn't option creep that broke 3.5, it was 'born that way.' ;P The surfeit of options provided for 3.5 deepened that system-mastery meta-game, and made rewards for it ever greater, but that was only a matter of degree (and not even that great a degree: in the end, 3 out of 4 of the Tier 1 classes were from the PH1). You could manage the brokenness with metrics like the Tier system, or variants like E6, or just put together a group with comparable levels of system mastery. 4e, conversely was neatly, even robustly balanced, and while the glut of options for it added a great deal of chaff, particularly to feats, and deepened the system mastery metagame, it didn't open up vastly greater rewards for system mastery, nor otherwise 'break' the game, just added to its complexity if you used it all (and, FWIWW, the on-line CB did manage some of that complexity for you).
Now the concern that 5e could be broken by a glut of options isn't unfounded, it's just not looking at the complete picture. 5e and 3.5 classes are balanced,
mechanically, in basically the same way, each has a list of classes with a long list of abilities (mostly spells) and varied resource-management schemes, and the relative power of those classes balance at some theoretical point, often summed up (though it's more complicated than that) as a certain number of encounters/day - in 5e, it's 6-8. Add new abilities to those lists, and you affect balance. Change the pacing of the campaign, and you affect balance. 3.5 (and, more importantly, it's player community) expected the DM play 'by the book,' so that balance was fairly brittle - deviate from the balance point or add too much material and it broke, even without the application of over-rewarded system mastery.
5e, OTOH, doesn't just have less material pressuring it, it Empowers the DM to deal with balance issues as they arise. The DM can always rule how things are resolved, as well as controlling the challenges & rewards extant in the imagined world, and being able to strongly influence (even force) pacing. His hands aren't tied by an expectation of adherence to the RAW or even of formally-introduced and consistently-applied house rules - he's free to rule in favor of fun, slapping down a 'broken' too-powerful element here or bolstering a lagging PC there, as needed. That's not quite unprecedented in D&D history - some DMs ran 1e that way - but 5e's really embraced it. What's more, if you're sticking to 3.x-style RAW, every bit of the game has to be balanced, not just in itself relative to comparable choices/resources, but it has to do so even if it synergizes with some other bit. But, if the DM is taking care of balance in-play, he only has to worry about the choices the players have actually made and the bits actually coming up in play - and that set won't grow just because there's new material added.
So 5e shouldn't run into any meaningfully more challenging balance issues as it adds material. Even if, hypothetically, it were to get new material at a rapid pace like 2e, 3.x/PF or early 4e did.