Well Hillary was projected to win the election based on polls but Yrump ended up winning because he out performed in certain states. These states had people come out to vote for Trump that were not protected. I attribute this to either the i investigaton that started before the election or the quite Trump supporters. Thats not to say that there were not quite Hillary supporters too in certain states because she was also very disliked. 

Even for my experience I know a lot of folks that support Trump lowkey but dont say that they support him because leftist tend to be more aggressive that right wng folks. Now in not saying this is universal but this is what ive personally seen.

I feel kike Trunps best case scenario would be winning PA and WI. Ouke even though the polls show what they show I think we gotta take history into account. Ohio, and Iowa went big for Trump last cycle,  Montana is a Republican state im terms of POTUS elections, the best Democrats have been able to do in Texas in recent memory was in 2018 and they still lost by 2.1%.

This is mostly anecdotal.

Again, I have to ask, if shy Trump supporters were a significant factor, why would we expect that to be limited to one small geographic region?

I put sort of a lot of time into looking at the data. What in the data suggests that there is a significant number of shy Trump voters? Or what data am I missing?

Well I don't have all the answers but for example WI Hillary was leading with over 7 points which is higher than the margin of error so thats a logical explanation. 

It's not a logical explanation, it's a hypothesis. It becomes a logical explanation when you test and verify it.

If the discrepancy in Wisconsin was the result of shy trump voters, then wouldn't you expect to see similar results across all the states we have data for? Yet, that's not what we see. Over half of the swing states were within a point, or actually overstated Trump support. And 13/16 states were within the margin of error.

Unless there is an explanation of why there would be so many shy Trump voters in the midwest, but virtually none anywhere else, the "shy Trump" hypothesis fails. 

I've been thinking here that, as long as Florida, Georgia, and North Carolina are called for Biden, the election could be called as soon as polls close in the West Coast, independently of whatever is happening in the Midwest, where votes will take much longer to count. Of course, by then, people would long be celebrating.

A hypothetical map at that point in the evening, with gray states yet to be called:

Texas should perhaps be gray there but it makes no difference to the argument, really.

Alternatively, Florida is tight and is called after the West Coast closes but before Arizona, Nevada, Minnesota, and Hawaii are all called. In this scenario, any of these five states could be the "tipping point".

There is probably a 0% chance Florida gets called on election night, maybe a 1% chance it gets called the next day.  Basically, don’t expect to know Florida for at least a week...

Only if it's very close. According to 538:

"Timing of results: Should be very fast. Florida is accustomed to handling a heavy volume of mail ballots and has laws (like letting counties process absentee ballots weeks in advance and not accepting most ballots that arrive after Election Day) that encourage an early count. In other words, results should be nearly complete within a couple hours of polls closing. That said, counties are allowed to take their time if needed, so some larger counties may still be tabulating mail ballots on Wednesday or later. That means that if a race is close, we might not know who won on election night."

Besides, it has a rather unique pattern of counting votes (tabulated ballots and early votes -> e-day votes -> late ballots) meaning if Republicans aren't leading at some point on Tuesday, they aren't taking it at all.

Florida should be called day of or Wednesday. They allow ballots to be processed like a week or more in advance, and don't accept ballots received after election day. Unless it's super close like 2000 or there's something weird like in 2000, it should get called quick. Which in turn will let us know if we're in for a long "election night".

Should and will are very different things.  Florida has a pretty extensive history of screwing up or close elections.  I’m just saying, I wouldn’t plan on Florida being an election night call.