Aristotle was not an empiricist in anything like the modern-day sense of the term. He had a "philosophy of science" radically incompatible with the empiricism of the early moderns or the mainstream philosophy of science today.
In particular, Aristotle held that science would consist in an axiomatic-deductive system, for which Euclidean geometry served as the exemplar. Along with this, his epistemological stance was that we know the theorems of the axiomatic-deductive system by virtue of grasping first principles and deducing those theorems.
Thus, Aristotle's philosophy of science consists in the search for first principles. This yields the traditional distinction between episteme and sofia and nous. Knowledge of first principles, which is yielded by nous, is more certain ("better known") than the episteme derived from the first principles. The ultimate first principles grasped by nous would be, for Aristotle, an a priori ontology that ultimately grounds every specific science.
But empiricists hold that we do not accept a scientific theory because we believe its axioms. Rather, we accept the theory because it yields a set of predictions superior to competing theories. We cannot then have ultimate knowledge of any first principles; we can only make strategic decisions that will refute a theory.
Empiricistic philosophy of science rejects the whole search for first principles and the possibility of an a priori ontology. The basic thrust of empiricism is that all scientific knowledge is hypothetical, and that the "first principles" of a science are as hypothetical as the theorems themselves. There is no problem of first principles in today's philosophy of science, rather, ontology becomes an a posteriori handmaiden of the empirical sciences.
It's true that Aristotle holds that knowledge of abstract objects is abstracted from empirical particulars. But I don't think Aristotle is anymore "down to earth" by today's standards than Plato. He had what is in my view a more sophisticated philosophy of science, but by today's standards it still looks exceedingly rationalistic. I don't mean to suggest that the rejection of Aristotlian ontology by empiricists is right or wrong, but merely to say that Aristotle's position on ontology and epistemology gives a far more robust place to mental faculties like nous and sofia for "seeing" mathematical forms than does present-day philosophy of science (rightly or wrongly).