For me, the crux of this situation has been mentioned briefly, but doesn't often get considered. Most people buy a product for "what it is" at the time of purchase, not for "what it might become." Only service industries can get away with such a strategy, but this would never fly in products - no one will buy a guitar without pickups or a car without tires because they "might come" in the next year. I happen to consider SONAR a product, and since "subscription" gets beaten down readily, others would confirm this.
The mindset of "product in hand" carries significant weight, and this was, in fact, the previous model. Folks knew what X3 had prior to purchase, then decided if they wanted it or not. For folks in a renewal situation they already have 100% at the
time of purchase and are paying for the "future," which may or may not contain anything users have requested. A large portion of what users are requesting languishes while things users never requested suddenly pop out of the woodwork, yet there is a lot of lip-service to the community having input.
One thing that occurred to me recently (and what ultimately bothers me) is that
SONAR is a host application first and foremost, and as such should be impeccable as a host (i.e., the things no VST can ever do for it). Functionality, usability, exception handling (big one), optimization, et al. should be the #1 priority above all else IMO. Everyone loves a good host.