There is so much to find fault with here that it would be funny, if it weren’t so outright dangerous.
First: You have no understanding of the different purposes, and therefore differences in methodology, of pre-election polling and exit polling. The very article you link to explains it, but you chose not to quote the part that undermines your conclusion:
When you see those adjustments made shortly after poll closing, that’s because we’ve gotten a whole lot more real information to tell us what the turnout is going to be by region…
Exit polls must be adjusted based on results, because the pollsters are trying to get an accurate read of the results by demographic, within districts, not overall. In contrast, pre-election polling is intended to accurately predict overall results (i.e. across districts and demographics). And to get accurate results by demographic within districts, you take larger samples from less populous demographic groups than you would for a pre-election poll. So until they start seeing actual turnout counts, the exit poll totals are meaningless.
Why would this matter so much in this particular election compared to the others you cite? Because there has been a huge rural-urban and demographic split in the results across just about every state in the Democratic primary! So it is in fact entirely consistent with the results that the unadjusted exit polls showed quite different outcomes from the final totals.
Second: Your conclusion about the impact of polling station issues is just wrong. From your article:
This consistent, unusual closure hints at electoral fraud, since Bernie Sanders’s base (younger voters) votes later in the day because they have school, jobs, and generally more responsibilities.
The problem is that some of the “closure” issues you cite were problems with polling stations opening late, not closing early. See this election day update from 538, for example — the station was supposed to open at 6:00 am, but as of 8:00 am voting was not yet underway.
(And by the way, some of us people under 40 with “jobs, and generally more responsibilities” actually vote early — I was at my polling site at 6:30 am. It’s a lot easier for some younger voters with jobs to vote early than late, so be careful about the conclusions you draw without evidence.)
Third: Everything you write about the insecurity of voting machines is pure innuendo. If you have a documented instance of actual hacking, please cite it.
Fourth: You take anecdotal evidence from your own community of like-minded individuals as conclusive of the demographics of who was affected. Again, see that 538 post: the people quoted were over 65. Just because young, white Sanders supports were the majority of people posting about these issues on the internet doesn’t mean they were the majority of people affected.
Fifth: The “purging” of voters happened in districts that went overall to Clinton, not Sanders. Despite what you may have heard on television, not all of Brooklyn is young and white, and not all young white people in Brooklyn voted for Sanders over Clinton — see for example this piece on Slate about a mostly white neighborhood in Brooklyn that went 60–40 to Clinton despite the author’s pre-election assumption that most of her neighbors supported Sanders.
I find it extremely troubling that you have put this forward as a rigorous analysis, and that so many people seem to agree with it despite the many and obvious analytical, investigative and logical deficiencies.
Please, try to approach the question without a pre-formed conclusion in hand, and with the important “razors” (Occam’s and Hanlon’s) firmly in the front of your mind.