my current best guess

FWIW, from Iris Murdoch's Metaphysics as a Guide to Morals (p. 323): "If we study one moral concept we soon see it as an aspect of another. It is true on the one hand that as moral agents we tend to specialise. The high-principled statesman may be a negligent father (and so on). It may seem as if we have a limited amount of good motivation available and cannot expect to be decent 'all round'. There are familiar ways of characterising people in terms of individual characteristics. Yet also a closer look may show this as superficial, and we then wish to say that the impulse toward goodness should stir the whole person."

Here 'tis: https://www.lesswrong.com/posts/iofy4cWC9AWzZDtxc/notes-on-gracefulness

I'd be curious to hear your thoughts about how insight-frisson might be induced by psychedelics/marijuana in terms of your model. Anecdotally, these drugs seem to promote both a lot of false-positive insight-frisson experiences (the feeling of having an insight is vividly there, but the insight itself seems to dissolve upon inspection) and genuinely insightful insight-frisson experiences (a conceptual discrepancy you didn't even realize you had suddenly comes to light, and a way of resolving it follows soon after, in a way that endures beyond the acute drug experience).

It sounds like you want to say things like "coherence and persistent similarity of structure in perceptions demonstrates that perceptions are representations of things external to the perceptions themselves" or "the idea that there is stuff out there seems the obvious explanation" or "explanations that work better than others are the best alternatives in the search for truth" and yet you also want to say "pish, philosophy is rubbish; I don't need to defend an opinion about realism or idealism or any of that nonsense". In fact what you're doing isn't some alternative to philosophy, but a variety of it.

A hypothesis that explains the perceptions can be a just-so story. For any set of perceptions ζ, there may be a vast number of hypotheses that explain those perceptions. How do you choose among them?

In other words, if f() and g() both explain ζ equally well, but are incompatible in all sorts of other ways for which you do not have perceptions to distinguish them, ζ may be "evidence for the hypothesis" f and ζ may be "evidence for the hypothesis" g, but ζ offers no help in determining whether f or g is truer. Consider e.g. f is idealism, g is realism, or some other incompatible metaphysical positions that start with our perceptions and speculate from there.

An author I read recently compared this obstinate coherence of our perceptions to a GUI. When I move my mouse pointer to a file, click, and drag that file into another folder, I'm doing something that has predictable results, and that is similar to other actions I've performed in the past, and that plays nicely with my intuitions about objects and motion and so forth. But it would be a mistake for me to then extrapolate from this and assume that somewhere on my hard drive or in my computer memory is a "file" which I have "dragged" "into" a "folder". My perceptions via the interface may have consistency and practical utility, but they are not themselves a reliable guide to the actual state of the world.

Obstinate coherence and persistent similarity of structure are intriguing but they are limited in how much they can explain by themselves.

It's a characteristic of philosophy, too, at least according to the positivists. If you're humoring a metaphysical theory that could not even in theory be confirmed or falsified by some possible observation, they suggest that you're really engaging in mythmaking or poetry or something, not philosophy.

This is a brief follow-up to my post “Redirecting one’s own taxes as an effective altruism method.” Since I wrote that post:

1. Scott Alexander boosted (not to be interpreted as endorsed) my post on Astral Codex Ten, which helped to give it more than typical reach.
2. In a flinchy spasm of post-SBF timidity, GiveWell explicitly told me they did not want to get their hands dirty with my donations of redirected taxes any more.
3. My tax arrears for 2013 ($5,932 original tax + ~$5,467 in interest & penalties) were annulled by the statute of limitations.
4. I made a $5,932 donation to Charity Entrepreneurship to celebrate.

According to Seigen Ishin (Ch'ing-yüan Wei-hsin):

"Before a man studies Zen, to him mountains are mountains and waters are waters; after he gets an insight into the truth of Zen through the instruction of a good master, mountains to him are not mountains and waters are not waters; but after this when he really attains to the abode of rest, mountains are once more mountains and waters are waters."

(D. T. Suzuki, Essays in Zen Buddhism, First Series, 1926, London; New York: Published for the Buddhist Society, London by Rider, p. 24.)

1. We inhabit this real material world, the one which we perceive all around us (and which somehow gives rise to perceptive and self-conscious beings like us).
2. Though not all of our perceptions conform to a real material world. We may be fooled by things like illusions or hallucinations or dreams that mimic perceptions of this world but are actually all in our minds.
3. Indeed if you examine your perceptions closely, you'll see that none of them actually give you representations of the material world, but merely reactions to it.
4. In fact, since the only evidence we have is of perceptions, the "material world" is more of a metaphysical hypothesis we use to explain patterns in our perceptions, not something we can vouch for as actually existing.
5. Since this hypothesis is untestable, it is best put aside when we consider what actually exists. The "material world" is not a thing, but a framework and vocabulary useful for discussing regularities in what is really real.
6. What is really "real" -- what the word "real" means -- is our moment to moment perceptions and interpretations, which appear to us in the form of a material world that we inhabit. 
7. GOTO 1

How to best break out of this loop?