I think the point was that even if you ask the customers, you end up with products that don't satisfy everyone, and you get edition wars. That's not something "going wrong". That's the nature of the beast - you cannot please everyone all the time.
On this note, I really, really recommend Malcolm Gladwell's talk on
what we can learn from spaghetti sauce.
Moral of the story: There is no one best D&D, just like there is no one best spaghetti sauce. Some people like regular spaghetti sauce, some people like spicy spaghetti sauce, and some people like extra chunky, and they often cannot articulate that fact. In all likelihood, similar "clusters" of preferences exist within the D&D community, and we're probably just as bad at articulating them
To continue with the pasta analogy, imagine Roberto's Awesome Pasta Sauce, which is a mildly spicy sauce with moderate chunks. Fans of both spicy sauces and chunky sauces like it. Roberto decides he needs a new recipe, and makes Roberto's NEW Awesome Pasta Sauce. However, this new sauce is a thin spicy sauce, with no chunks. Suddenly the chunky sauce fans feel "left out," or "disenfranchised," or "fired from the target audience." However, lots of new fans who DIDN'T like chunks are also brought into the brand.
---------
Without defining them, I'd warrant that D&D cluster preferences line up somehow with popular "styles" of game, even if they're not necessarily the ones we would rattle off from introspection. The move from 3e to 4e represents a change in which styles are considered primary/encouraged in the core game. Some ones that were previously encouraged are left out, and some previously unaddressed styles are brought in.
There's nothing inherently good or bad about this; what matters is whether it
works. There's a few key assumptions involved in this strategy:
1) The gain in marketshare/income/market growth/whatever from adding new styles will be enough to compensate for losing the old ones. This applies both in the short term and in the long term. With limited data, it appears to me that 4e has succeeded at this in the short term (lots of core rulebooks sold), but may have trouble in the longterm (decent number of people going back to 3e, doubt about the longterm investment of new players).
2) The ill-will generated by dropping certain styles won't have a significant impact on future marketshare/income/market growth/whatever. IMO, this is the one where 4e has done poorly. The GSL debacle created rallying points (former 3PPs) for the left-out clusters. I would hypothesize that a more lenient GSL that brought in more of the 3PPs would have have prevented the "opposition" from materializing as much as it had. A lot of non-4e-fans might have come around to it through house-ruled versions put out by 3PPs. (Fourth edition rules, third edition feel?)
