I agree! I think there's a significant market for straight UTAU-ports, if done well. If Teto UTAU's samples were made available as a concatenative bank in SynthV for example, it would cause an absolute storm. Breaking into that market is probably the most astute decision Yamaha has made.
I've probably mentioned this before, but there was a big Japanese poll asking whether people preferred Teto UTAU or Teto SynthV - and UTAU won by a landslide. This suggests that SynthV is getting more usage because it's just so much easier that it outweighs price and tone preference. (But as a caveat, I heard UTAU has a stricter licence, which would have some influence too - I'm not sure how much.)
However, a good UTAU port is easier said than done. UTAU has a lot of depth, and attempts like SV1's Renri didn't match the original. To fully meet people's needs, it would also have to be compatible with crosslang and autotune, as these are the two major usability advantages of AI. That's got to be a technical challenge, one I'm curious how Adachi AI will tackle.
We're hitting the cusp of the flaws we fought to remove becoming stylistically desirable, as happens in all art forms. Adachi Rei is proof that a commercial engine who invests in "retro" robotic vocals is a smart one!
Yamaha hasn't done much to impress me on the technical front this era, I'm interested to see if they'll pull this off. Too much additional buzziness and Rei will lose her signature tone (not the type of roboticness you want), fail to match UTAU's customisability and power users won't migrate. There are many things that can go wrong, but it's a goldmine.