I'm glad that's the point you take away :/
A surprising number of voters switch party from one election to the next, so the number who have "ever voted Tory" is going to be a lot higher than their highest recent vote shares. The survey below suggests a third changing their vote is commonplace these days, and it was an astonishing 43% between 2010 and 2015 (skewed by the Lib Dem collapse).
The British electorate is volatile but one party has to benefit more than the others for that to actually affect the outcome.
theconversation.com
Bear in mind, that's just between general elections - obviously a lot of people would have voted differently in European and local elections, further increasing the pool of of people who have "ever voted Tory".
Obviously turnout isn't close to 100%, but I'm not aware of any research which suggests non-voters wouldn't vote in a similar way, and given that there's considerable churn amongst voters/non-voters from election to election, the number who have voted at some point is a lot higher than the turnout for any single election. For example the vote in 2017 was just under 70%, and only a couple of points up on 2015, yet 17% of 2015 voters didn't vote in 2017. That's just voter churn between two elections two years apart.