For me, the definition "A group is a set of transformations on an object such that..." is so much more enlightening than "A group is a set G together with a binary operation * such that..."
Only yesterday I was trying to learn about differential forms. Most of the notes I found online introduce the wedge product in a deeply unhelpful way, by listing some axioms that it satisfies and deducing results from the axioms. It takes hours of work to understand why those specific axioms were chosen. For me - and maybe I'm wrong here - that's Bourbaki's formalist approach in a nutshell.
It was only when I found Terry Tao's notes [0] and Dan Piponi's notes [1] that I could actually see the use of differential forms. It's an unfortunate state of affairs for the discipline that in order to learn about X, you have to google "X intuition", since it's not given to you as a matter of course.
[0] https://terrytao.wordpress.com/2007/12/25/pcm-article-differ... [1] https://github.com/dpiponi/forms/raw/master/forms.pdf
so it's always been irritating to me how much introductory or "applied" texts bias themselves towards physics and engineering. like how a lot of multivariate calculus/introductory real analysis texts pretty much limit themselves exclusively to three dimensions and teach notation to match.
it even plagues more advanced books; i'm struggling through Villani's text on optimal transport and nearly every example is from physics even though the theory was developed for economics and has extremely broad applications across many disciplines. half of the "motivating examples" i can't even make sense of what they're talking about.
as much as I struggle with the hyper-abstract bourbaki-style stuff, at least I have a hope of understanding it without needing to dip into an irrelevant discipline where i have no background or historical context.
And for example, a real revelation with regard to infinity came, when I read (or heard somewhere, it was long time ago) that infinity can sometimes be few millimeters or even micrometers. Up until then I always imagined some really large number, but at that point I realized, you need to put problem before you need to put problem before you into perspective.
At times he went a bit overboard, but he makes some valid points in this lecture: https://www.uni-muenster.de/Physik.TP/~munsteg/arnold.html