What I don't understand is why WOTC is getting slammed for 'large errata' when Paizo are doing the same thing, in the same context of the title of this thread? Paizo's customer service is fine, I've bought their products and will continue to do so, but the errata is still there. There are problems with the way WOTC has handled D&D, but errata is not one of them.
Errata is a big part of it actually, or at least the expectation that something has to be so finely tuned that even the smallest error needs to be corrected immediately to avoid severe problems. Paizo may make the same amount as WotC in the end, but how they get there is completely different. They tend to make lots of small tweaks that can be measured more accurately, letting the changes build up over time. WotC has an annoying habit of simply throwing out the entire system and trying to completely rewrite it again and again and again. This isn't just true of 4E, it was already very evident in 3.5. Completely rewriting polymorph 3 times rather than trying to isolate the actual problem and deal with it while leaving the rest alone was completely silly, and that's how WotC tends to approach errata and rules changes in general. Paizo adapted a fair number of things initially, but since then has been more incremental in their overall approach, making it easier to adjust and adapt.
As for Paizo having the advantage of a well tested rule set, that is precisely part of my point. They didn't throw the baby out with the bath water, so to speak, but rather found ways to tweak and improve on the existing baseline within measurable bounds, making their job a lot easier, both in satisfying their customers and in being able to effectively measure changes. They didn't change everything they could possibly change just for the sake of change, making further changes and future expectations much more realistic to accomplish. Even with the new classes and archetypes and whatnot, they made it very clear early on that the new options were not automatic assumptions in any world; they also frontloaded most of the player options upfront precisely in order to get that part out of the way for a year or two so that further changes could be made with a fair bit of data on hand. I have no doubt that when Pathfinder 2.0 comes out, they will find ways to incorporate what they have learned about the new base classes and archetypes into the existing core rather than try to dramatically rewrite much of anything from scratch, just like they have started to incorporate material from later core books into Golarion as they get more data on what works and what doesn't. And at least so far, price has not been a major issue with the constant updating of pdfs for free after initial purchase. Initially WotC did this as well with 3rd edition being in many cases simply universally accepted houserules. The release of 3rd edition didn't cause the same amount of split that 4E did because even if details were different, even most of those were at least familiar in concept to existing mechanics and popular house rules. Even 3.5 didn't attempt to reinvent the basic wheel for the most part; it was the cumulative number of small changes that set apart 3.5, not any one big change. Both did really well because the bumps that were there, for the most part, were well documented, and easy enough to work around or simply ignore that it didn't bother most people enough to stop playing the game overall. With 4E they didn't bother to establish a baseline to make changes from, and that makes making any kind of effective errata or rule changes next to impossible without creating a lot of additional problems; that's something they have to avoid with Next. It's what drove a lot of people away from both later 3.5 (where the splat books just plain started getting ridiculous) and 4E.
I guess at the end of the day, I personally find Paizo's approach much less of a headache than wondering how WotC is going to stick their foot in their mouth for the umpteenth time next. I'm not anti-WotC, otherwise I wouldn't be bothering with paying attention to Next at all, but I do need to see them learn from their past mistakes before I give them the same benefit of the doubt that I give Paizo, which has earned it over time. If they can do that, than how often and what form the errata takes isn't going to matter; if they can't, it's going to be another nail in the coffin, just as it was for 4E. Essentials proved to me that is was possible, now they just need to keep building on that.
To be fair, the problem isn't entirely WotC itself; both their fanbase and their distribution model has changed, with both becoming more reliant on official errata that keeps up with the speed of the internet. The internet will always break things much faster than WotC can keep up, and demand fixes at an equal speed while doing so, something that no company could keep up with. Their decision to largely abandon books and printed magazines for an almost entirely digital format put them in a tight spot that they have worked with as well as can be expected. I will never try to argue that another company in that same spot could exceed what WotC has done. Paizo definitely has an easier road here, sticking with pdfs and maintaining for the most part the same, far more reasonable expectations of corrections and changes that come with a book format; while the SRD may be the most used tool at the actual table, pdfs and physical books still remain vitally important, making expectations based on those mediums much easier to maintain. Their strength is in learning from others mistakes, something that I will freely admit that they have a much easier time at doing than WotC right now, but as Paizo continues to grow, and WotC faces an uphill battle to sustain their new edition while also maintaining 4E as a semi-active system, at least on DDI, that will change, and the true test will be to see if WotC can learn and adapt to the changing conditions as quickly as Paizo has and probably will for the immediate future.