If Remain had won 52% to 48%, you can bet that no second confirmatory vote would have been suggested. Can you think of an example where the pro-EU side won but the result was so close that it was felt a a second vote was needed? That is my point.
EU treaties were voted against by referendum in Denmark (1992) and Ireland (2001 and 2008). There were second votes where the treaty was subsequently passed in all three cases.
You can bet that it would have: https://www.theguardian.com/politics/2019/dec/14/dominic-cum...
How many of the elections you're referring to were contested, called illegitimate or had similar tactics used to effectively call for the same thing? Maybe you can show me that there are a bunch of examples without anything like this, but the way this tactic spread throughout the political right all over the globe in the last ~10 years makes the assumption quite reasonable IMO.
The examples I gave in Denmark and Ireland were of real re-runs of ballots organised by the state where the outcome was changed. Beyond the treaties of Nice, Lisbon and Maastricht I could have also mentioned examples in France and the Netherlands where the pro-EU lobby had a second bite of the democratic cherry by being able to make amendments to treaty proposals before organising a second ballot relatively close to the first. This is not something that usually happens after a referendum, which is why people link the EU with “keep voting until you get the right answer”, which I personally think is profoundly anti-democratic.
Either way, of course you won't find counter-examples if you conveniently define things so the requests made by the anti-EU side simply don't count.