Where do you think we are now?
Now that question can be interpreted in different ways. How many users there are. Or "subpar render engine". Or both.
We can only speculate about how many Cheetah users there are, but probably you're right, not that many.
Falcon is in several ways an astonishing renderer. It just has a few can't dos and several restrictions, the worst of it that it can't handle sss (also more refraction rays really would help). That all those restrictions still exist has certainly to do with Martin's plans all rendered mute through Apple which threw a lot of work at him. We simply don't know what would be there and what not. Even so I wouldn't call it "subpar".
And we do both agree that Cheetah simply can't deliver the same production quality of Blender. But being completely honest, Cheetah's modeler, even with a lot of work involved, couldn't compete with what's possible in Blender (and if you look at the available plugins and what they can do, it outshines some of the pricey professional tools also. In addition there is a strong modeler out there, Silo, which is good software (albeit badly handled by neverwinter, who lost a lot of trust of their customers. Even they included a renderer now because there was obviously some demand). As a more specialized tool (modeling) the customer base would change and it had to deliver more to have some success (that it's only available for mac would be a serious hindrance, too, by the way). So I'm not so sure if your idea of changing Cheetah into a modeler really would be very successful. Im sincerely not convinced. There is probably a reason why Silo for a long time wasn't updated anymore and apps like Hexagon and others more or less died (Hexagon wasn't bad by the way, a lot of promise with rather simple use, and is probably, even being some 15 years old, almost without update, still the better modeler than Cheetah's. It died over at DAZ, probably because there wasn't enough money in it and for quite a time available for free (I'm not sure if it still runs on Win, for Macs it probably doesn't even run on Big Sur)).
Now, just for the sake of argument, lets say that your business model really would make sense and as a newly overhauled modeler Cheetah would be a raging success among mac users (alone that restriction weighs heavily). But that would also mean that a lot of actual Cheetah users would be left behind, imho. Martin, in my opinion, is not in it for the money, but seems to really care about his user base, especially for long time users, and it's probably more important to him to give them what they want (a better Cheetah with a renderer) than raking in more earnings. Also, in my book, as a somehow restricted but "full" app, Cheetah stays a good solution for learning purposes.
If it was only about the money, then both wouldn't make much sense, Cheetah as it is, Cheetah as a modeler alike. With far less effort he could create and sell some plugins for Blender and / or other 3d apps and stop the development of Cheetah all-together. Financially that would make the most sense, at least in the world of 3d, for a single programmer with restricted time and almost non-existing PR (blender-market would help a lot finding customers). Obviously that's not what Martin wants. And I deeply respect him for that (besides being a good developer).