I find that this kind of talk misses the point. When I paint a person I'm painting a person, not painting paint, and when I write about a battle I'm writing about a battle, not writing about words. So when I see an apple I'm seeing an apple, not seeing an experience. But that doesn't address the epistemological problem of perception. What is the relationship between the paint and the person? What is the relationship between the words and the battle? What is the relationship between experience and the apple? What does it mean for the former in each case to be about the latter in each case, and to what extent is any information given in the former a product of that medium rather than a true, independent, property of its subject?
I brought up blindsight earlier. The body responds to external stimuli in a manner that lacks conscious awareness. What the direct/indirect realist wants to know is the extent to which visual percepts (that thing that's missing in cases of blindsight) "resembles" the external world object that is said to be the object of perception. Simply saying that the external world object is the object of perception or that experience just is the stimulus-response event (one or both of which you and unenlightened seem to be saying) doesn't address this question at all. — Michael
First, "we see room furniture, not head furniture" might not address the point you're interested in here, but it addresses Marchesk's point that what we know of the mechanisms of perception make it impossible that we see only room furniture and not head furniture.
Otherwise, maybe I'm not even interested in the question of how what we see "resembles" the external world. In fact I don't really know what that means. Or rather, I think it's a bad question.
214. What prevents me from supposing that this table either vanishes or alters its shape and colour when on one is observing it, and then when someone looks at it again changes back to its old condition? — “But who is going to suppose such a thing?” — one would feel like saying.
215. Here we see that the idea of 'agreement with reality’ does not have any clear application. — Wittgenstein, On Certainty
Asking how much our perception resembles reality, or gives us information about it, is akin in this context to asking, "what do tables look like, independently of how they look".
The question as to how much the appearance of things is a product of the perceptual medium presumes the possibility of appearance without perception. What you call a medium is what I call the stuff and processes and behaviours that
constitute perception.
Answering the question as to how much information we get about things through perception more charitably, I might say things like... quite a lot, it depends, often as much as we need, etc. I don't think this has much to do with the big problem that you see. We don't get much information about the shape of a building without walking around to the back.
But to get to what you're interested in and state my positive position more explicitly: we always perceive under an aspect. We perceive affordances, what is relevant. Perception is a coupling with the environment in ways that depend on perceiver and environment. This might be a form of correlationism and so not as realist as you'd expect, but in the same way that Kant didn't think of himself as an idealist, neither do I.