I would guess it’s probably happened before sometime in the 1800s, but I don’t know that for a fact. I also think you’re overstating things a little bit too. I don’t think that a significant number of Trump electors are really in play or ever were.
I recall reading here a month or so ago about the election at the end of the Grant administration. There was not a “clear winner” so a deal was worked out to compromise, end Reconstruction and bring the troops home (from the South) and Hayes could be sworn in. Smoke filled back room deal.
https://en.wikipedia.org/wiki/Faithless_elector
1960 - Republican elector from Oklahoma tried to deal with the Dixiecrats to elect Harry F. Byrd instead of Kennedy.
The other cases appear to be independent decisions by the elector.
The 1796 story is clearly incomplete. The vote was Jefferson 73, Burr 73, Adams 65, Pinckney 64, Jay 1.
If Miles was faithless, there must have been at least one other faithless elector from Pennsylvania (total EV in PA was Jefferson 8- Burr 8- Adams 7- Pinckney 7- Jay 0).
Also there was a “faithless” elector from Rhode Island who voted for Jay.
Had the final vote been Jefferson 73, Burr 72, with one of the Burr electors voting for someone else, then the presidential election would never have gone to the House at all. (Jefferson would have won the presidency outright, and Burr would have won the vice presidency outright).
The 12th amendment rectified the necessity of a designated faithless elector.
Oops....I was looking at 1800. They say “0 but almost 1.”
Although had Lispenard wanted Burr for President he didn’t have to vote for Burr twice - he could have voted for someone other than Jefferson - it would have then been Burr 73, Jefferson 72. Burr would have been President and Jefferson would have been vice President.
1796 - they say 19...which included one Federalist elector voting for Jefferson and 18 others voting for someone other than Pinckney. The actual vote appears much more complicated than that.