I suggest that some of the ftp numbers being quoted are on the high side.
In the Cycling Weekly Zwift 10 miles the winners are in the 300-350 Watts for 20-25 minutes men) , and in the 150-250 area for women. Given that ftp is a 1 hour figure, I’d expect somewhat lower figures for an hour.
These are riders who also figure in real world results, so they’re in the top percentiles of riders who race.
Having experience of running ftp and MMP ( maximum minute power) tests on a range of cyclists from relative beginners to elite level I’d say that the average, mid years, regular cyclist but not race trained would have a real world ftp ( male) of 150 to 250. That’s on calibrated equipment. We regularly saw people coming in with bigger ftp figures from bike-mounted power meters or from trainers, but they didn’t hold up to calibrated equipment scrutiny.
As QG infers, pulse is, in my view, a more reliable gauge for an extended effort. It’s A measure of stress on the “ engine”.
Terrain is only one part of the story, but for an Audax type effort it will be important. If the aim is to get round in a time limit it makes sense to reduce high levels of stress that will take you into anaerobic energy production. Appropriate gearing will be one part of this. In an ideal situation a rider could ride in zone 3 of the 1-5+ scale up hill and down dale, and be able to hold that for extended periods, as long as gearing is appropriate.