A couple thoughts:
1. I said when Trump was first elected that, while I was very concerned about his right-wing agenda, his policies were third on my list of things that were freaking me out. The first two (I flip-flop on which goes first and which goes second) were 1) the hollowing out of our democracy and 2) getting into a huge crisis that he was too incompetent to handle. If you ask me today, No. 1 is still in the lead, but the Soleimani assassination was a reminder that No. 2 is always lurking as a possibility. All of which is to say if there were an administration that took 1 and 2 off the table but 3 was still out there, I would sleep much better at night, but no, I would not vote for them, because I still view No. 3 as a BFD.
2. I think we tend to fall into the "all else being equal" fallacy. Back in the '80s, the science writer Douglas Hofstadter wrote a piece on
counterfactuals in which he pointed out that, at the time, some people were saying that if Jesse Jackson were white, he'd be elected president, while others were saying that, if he were white, he couldn't get elected dogcatcher. Hofstadter's point was that, when people suggest counterfactuals such as those, they're doing a "light mapping" in which they tweak one variable and use that to illustrate a point. So the pro-Jackson people were saying, "If there were a viable presidential candidate who didn't have to deal with racism, he could get elected." The anti-JJers were saying, 'If a white candidate acted as outrageously as Jackson did, he'd be finished."
I say all this not to crap all over the OP's premise but rather to point out that such counterfactuals are surprisingly complex, and all they really tell us is what other variables people consider most important. If you take them seriously and start to really dig into how they would play out, it soon becomes impossible to accurately analyze the situation. What would a Rubio Administration look like? How did he manage to get nominated by the type of GOP electorate that chose Trump? Is it really possible to isolate out the tweeting, name calling and corruption from who Trump is? Those questions aren't unanswerable per se, but any answer will necessarily reflect the worldview of the respondent more than a dispassionate analysis.
Of course, there are cases where it is possible to more easily isolate variables. Consider the counterfactual, "What if James Comey had never announced the reopening of the Hillary email investigation in Oct. 2016?" It's easy to imagine a world where that one thing doesn't happen, and so it's relatively easy to discuss the implications. (Nate Silver literally came up with a
quantifiable measurement of the impact of Comey's announcement). Whereas a counterfactual that assumes a non-crazy version of Trump (or a white Jesse Jackson) requires us to fill in lots of other gaps, which in turn introduces lots more opportunities for our own biases to affect our answer. Think of it like the difference between the refs blowing a call in the first quarter of a game vs. missing the DPI against Robey-Coleman at the end of the NFCCG.