And we wouldn't have a tenth of the flavor and detail in those settings if they had. TSR could do setting support, something WotC isn't doing remotely near the levels of 2e or even a fraction of what we had for 3e.
Meh. I'm not convinced there was enough good stuff in the post-release materials to justify the havoc wrought upon the settings - havoc which would then persist into future editions and re-releases, as TSR/WotC felt it necessary to preserve the setting's established lore, however crappy that lore might be.
I also think setting detail is overrated. In my opinion, a good D&D setting should have giant unexplored areas on the map, and grand mysteries that are never "canonically" resolved, so that DMs can fill in those blanks with their own ideas. Every time a new release fills in some blanks, it narrows the scope of the setting; it also creates a headache for DMs running established campaigns in that setting, since some of the new material may clash with what the DM has created.
Novel tie-ins, if treated as "canonical," are the worst, because fantasy fiction has a long tradition of calamitous events and world-altering plotlines. Which is fine... in a novel. But any D&D setting with novel tie-ins, unless the tie-ins are rigidly controlled, is going to be torn apart and rebuilt - repeatedly, if the tie-ins keep going long enough. Dragonlance is the poster child for this problem, but as Nymrohd pointed out, TSR did it to pretty much every setting they published.
Each of those cataclysms is likely to a) trash the setting's original concept, and b) sever ties with the setting's existing fanbase, since "their" versions of the world - the ones their campaigns are set in - are now fundamentally incompatible with the "official" version on which all future releases will be based.
...Hmm. Now I'm thinking about ways to avoid these problems while still providing support for a popular setting...