Why would a reset help here? the performance difference is vey clear when switching back /forth, this has nothing to do with LAN / wifi performance, it's the WAN side that's slower. maybe the test involves multiple resolve requests, I don't know, but it severely affects the test. just go to speedtest.net and see for yourself.Of course the DNS inquiry is slower (higher latency than providers DNS) over DoT, but it can not affect the speed when the address gets resolved!
Reset to factory defaults...
Once resolved, for sustained data rates, it should not matter, but appearantly that's not the way they test and also not what reflects real usage loading and refreshing different pages. So my conclusion is that for real life use scenarios your mileage may vary.