often get 90%+ participation. For example, just last month Putin has obviously orchestrated a non-binding independence referendum in Venice: 89% of participants voted for independence with
That's actually a very insightful (and it turns out informative) refutation.
You see, it turns out that Venice poll was not actually a proper plebiscite, but rather an online poll. Online polls aren't exactly famous for being accurate representations of popular opinion. I can't find any official number for participation, but it looks like they were "expecting" no more than half the electorate to vote.
So what would they have found in a real vote? Actual opinion polls on the question were apparently finding about 2/3 support for the idea. A 66% vote for a popular measure in a Democracy would be perfectly in line with what one would expect to see in a real vote.
But maybe you're right, and I should look at a real independence referendum. They are preparing to have one of those in Scotland soon, right? What are the predictions for numbers there? Well, current opinion polls seem to show the electorate running at about 50% to 33%, with the rest undecided. So we can expect that when election day happens, whichever side wins, it won't be by larger than 77%, and likely far less.
Or how about we look at actual numbers from an actual vote? Puerto Rico had a referendum on statehood/independence/etc. again two years ago. 78% of those eligible voted (so indeed high by western standards, but not ridiculously so), and the results were about 54% to 46%.
So I may have to grant you that turnout numbers may be a bit higher than normal during a referendum like this, but that's still worlds away from making the "results" the Russians reported from Crimea look anything like a real vote.