Maybe, maybe not. But it's pretty clear that an economy or society fits the definition of a complex adaptive system, and it's pretty common knowledge in that field that it's possible for there to be phase change transitions, where effects come suddenly and all at once at a specific level of participation of the independent actors.
Analogies include things like herd immunity, or sensitive dependence on initial conditions in weather forecasting.
Firstly, even a smallish percentage of immunity would still see population wide benefits.
Secondly, as diseases are transmitted to people physically nearby, this could be tested by immunizing a large percentage of a single population, such as a small town full of people who don't travel much.
If you are truly concerned about the population wide effects of UBI, then you could do something similar, like fund a small poor town, and give EVERYONE in that small town UBI.
Unfortunately, I expect that of such a test were attempted, you'd see the same exact criticism of it, though. It'd probably be something like "you can't see the benefits unless the entire WORLD is on UBI!!!" or some other further doubling down on the unprovability of it all.
My point being, that this scepticism of actually testing the damn thing seems more like a defense mechanism against the possibility of being wrong. It is just the same lazy approach that everyone has for their pet theories, as a way of deflecting any and all criticism by declaring it unfalsefiable
If UBI is so great, then similar "almost UBI" proposals should ALSO show good results.
I sincerely doubt that there is a magic line between "similar but not UBI" programs and "UBI" programs where the good stuff immediately appears.
If UBI is impossible to test, then we shouldn't do it.
It would be insanely risky to spend trillions of dollars on something that its proponents claim can't be tested. So we should just not implement it in the first place.