Are there any stats I could mine for this (and other) information? I want to get an idea of how much winners go all in, compared with for example me.
It depends on the size of the tournament, but the percentage of tournaments won without the winner ever being all-in is very, very low (<.05%).
Two things at play here: first, other players can choose to put you all-in, and if you're +EV, you've got to call. If you choose -not- to make +EV all-ins, because it "risks your tournament life" or something asinine like that, you're the fish at the table. Second, its exceptionally rare to have a tournament where you're the deepest stack all game. To get there, you'd have had to risk a SIGNIFICANT portion of your chips on numerous occasions in order to amass your stack. Which is a catch-22, because this indicates that you were already all-in, or close to it.
Winning a tournament without going all-in is a pipe dream and is the wrong thing to try to fix if you're trying to win more money at tournaments.
Source: A few million tournament hands.
Other than mining published tournament data or summaries of your own data sets, I don't believe there are direct avenues to access this specific kind of data. That said I'd be willing to bet, with almost absolute certainty, that there isn't a tourney in history (with significant participants) where the winner wasn't all in at some point prior to the final hand.
That said, the frequency of their all ins is tempered by the stack size they posses. Better players = larger average stack size so it probably less than average frequency, but considering chip-to-stack ratio it's still probably all in multiple times per tournament.
Equity per decision is probably more relevant.