Carnap and the Meaninglessness of Metaphysics The entities that are investigated in (most of at least) the sciences are phenomenally real enough for us — Janus
I wouldn't say so in the sciences, soft or hard; we don't have 'intuitions' of things like populations, physical forces, and so on.
What we come closest to having 'intuitions' of are tangible ordinary objects, but even these are thought of transcendentally: we project them as seen from 'infinite sides,' and we never have an intuition of their totality. So, treating things as 'objects' is itself just a regulative idea. Unlike with metaphysics, though, it's a practice of using regulative ideas that tends to do useful work in daily life, likely because our language and cognitive faculties are adapted to do so, whereas the kind of metaphysics philosophers do was invented a couple thousand years ago, as the result of leisure time leading to funny linguistic puzzles (essentially, philosophy proper begins in sophism), with no native practical application.
The sciences are therefore almost entirely hail marys linked to employing these transcendental illusions, but even there, we do get some effects out of them (mostly technological, though we have no way of really controlling or even understanding its effects). Ordinary life is a bit closer to home, but even there, we have to act as if we cognize things we don't to get by, and life is basically a bunch of regulative pretenses that justify what we do.