[strogonoff]

Do you choose one model over the other if your basis for it is along the lines of “it sounds weird and unfamiliar to me” or “it sounds less like things I deal with in daily life” or “I cannot imagine what it looks like”? Should you do it?

Surprisingly, in some ways, yes. A model is ultimately an attempt to lossily explain something in terms of (map it to) something else, more familiar. If it fails to do that, maybe it is not a good model.

However, you cannot stop at the term. Yes, we all have seen a black hole (hell, I have one in my bathroom) and none of us have seen IRL a multi-dimensional anything, much less a supermaze fuzzball—but if you consider what the black hole model actually is, with things like event horizon, Hawking radiation, etc., it quickly loses its intuitiveness.

Finally, of course, I’m not a string theorist, but I reckon if your model for something includes singularity as one of its main features then you obviously need some better model.