At various points, I have been accused of treating too harshly with Abstractions, in saying that every level of abstraction is a level further away from reality, such that e.g. pure math is the most unreal of all.
I have also been accused (gasp!) of actually relying on abstractions, even though I grade them as unreal.
To the latter, I readily agree: I am with the quasi-realists in saying that we don't have a choice. My degradation of the abstract is for my own purposes, for mental hygiene, which leads to the first accusation, and an answer to it:
Keeping in mind that abstractions are less real than the data is was derived from can help us notice whem very competent people make very silly mistakes.
Case in point: Physicists waxing poetic about how if some constant were just a little different [scrreeeech] let me stop you right there!
Reality doesn't have constants.
Constants are numbers (often with units attached) that we stick into our formulae to make them work.
But the formulae are not reality.
We have absolutely no reason to think reality has constants.
Hence, the only way a constant could be different would be if reality was different (this happens a lot: reality turns out to be slightly different from what was previously believed, thereby correcting a constant ever so slightly).
Reality itself, we find again and again, doesn't change.
Every day we notice that, yeah, still the same.
Because to all intents and purposes it seems to just be spontaneously trundling along the only way it can.
The map is not the territory. Physics is not reality. Reality is not made of math, just because physics is.
Mental floss, see?