+5
Seems dissonant to say the polls are wrong because Trump supporters are being quiet, but then also cite a poll as evidence of it. That's really a narrow thread, to saying they're answering general polls but not specific ones. Also, pollsters weight to account for these things. It's imperfect, but I think a lot of laypeople think they're just calling whoever and just shrug it off if a certain demographic decides not to answer. That's not how it works.
Anyway, what you're describing has been described before, it was once known as the "shy Tory" effect and it's been floated with different names and in different contexts in many elections since. And it'll be cited again after this one even if it doesn't end up manifesting. I'm pretty sympathetic to the idea that it exists, but thinking it can explain away the kinds of margins we're seeing is a pretty huge stretch.
If the results are massively different than expected, it won't be because of this effect, it'll be because of a more fundamental problem with modern polling.