> it's just that nobody would invest in training publicly usable models if they could be easily distilled.

Thats literally what is happening right now though. People are spending hundreds of millions on a training run, and then people are distilling them, fairly easily, and making cost competitive models.

We are seeing all of this in action right now.