I don't see this happening as long as they have the official beta testing crammed full of every day users who wouldn't say "the well known substance" (even with a mouth full of it) about the release candidate's shortcomings for fear of not being invited back for a return gig the next year.
The flaws are far deeper and more wide-spread than that.
From what I can tell, the beta testing period is a brief timeframe during which they are simply trying to identify new quirks introduced by recent code changes, and get some of those fixed before everything goes to production. Even if they get feedback at that point that an entire new feature was designed incorrectly from the ground up, it's too late at that point for them to do anything about it, and they aren't interested in that sort of feedback, even though it's really the most important thing they could hear. It simply doesn't do them any good in the beta test phase; it's already too late. And once something wrong has been introduced to the product (e.g., Siteless Alignments, the horrid design of Parcels, the horrid design of Labels, etc.), it tends to stay in next year's version too, even if they get feedback in the beta period that they should have done something else.
But this is only one of many, many, many ways Autodesk's software processes seem to be flawed. It's difficult to say what's happening inside of Autodesk, but some things shine through. Just from what we can see, Autodesk's software processes are seriously flawed, starting with the way marketing drives development, and going from there. And it's not really like they can do something simple, like simply stop the yearly release cycle. The yearly release cycle in itself is not the reason we see so many bugs, the bugs are the result of other aspects of the overall development process, which includes everything from the definition of new requirements to the QA/QC. So many things are wrong that they can't simply make a few tweaks here and there and fix it - they need systemic change, and that's something that is very difficult to accomplish in a multi-national corporation like Autodesk. It doesn't help that these changes also cost money in the short term, which has a strong tendency to blind those in charge. Changes that bear significant fruit in the long-term but incur a short-term cost are often ignored, especially when stock holders are involved.
So here we are, here we've been for years, and here we'll probably stay. Autodesk has no real incentive at this point to change anything. Right now, all they have to do is expend
just enough effort that their existing customer base doesn't migrate en masse to other software, which is very easy to do, since sane people don't want to deal with the hassles of evaluating and purchasing new software, then retraining staff, reworking processes, dealing with legacy projects that were done in the old software, etc. And of course, most people are lazy. I'm lazy myself, which is why I wrote the Sincpac-C3D.
