The challenge is that all the paths are difficult and risky, so it's always possible to find reasons why Hard and Rocky Path A is hard and rocky. But the alternative is Hard and Rocky Plan B and it's easy to show why it's hard and rocky too, so we have to settle for a plan that has risks (for example, for Levit's votech model, top professors want to do research and get paid well so if research isn't valued then we'll end up with 2nd or lower tier professors whose teaching wouldn't be that much better than smart grad students, and the non-research model is lower profit, which means lower paid professors and more pressure for lower quality and heavy reliance on adjuncts, who, well, get paid even less than TAs! And will often be worse teachers than a smart TA, you'll have adjuncts teaching at TU on MWF and TCC on TTh, which sets up a death spiral where it's harder and harder to deliver value to students and thus harder and harder to get them to pay a lot, until you're charging the same as OSU but without state support, and so on).
Whatever the path, we need to be clear about the goal. People say "we can't compete with MIT and Stanford on research so we will fail". But will we? That's like "if we can't beat Alabama most years, we might as well drop football". But is there a more modest vision of success that is realistic and more achievable, that we all agree is valuable and sustainable? Maybe 15% of MIT, making up a number, but still worth doing?
And as the always astute boobay says, maybe focus initially on lower cost research, not nuclear fusion where you need a half mile linear accelerator or whatever. Do you know how much has been invested in educational tech (ed Tech) and health and wellness tech companies in the last 5 years? An astronomical amount, think tens of billions, with a "B". And those can come out of dpts with modest needs like psychology, teaching and nursing, as long as there's a strong computer science program. I mention those only b/c I know about them but I'm sure there are plenty like that.