I find the relative performance chart at
https://www.techpowerup.com/gpu-specs/geforce-gtx-1080.c2839 to be useful for estimating GPU performance when considering whether a GPU is going to be satisfactory.
1070 is 85% fps of a 1080 --> okay for 60 Hz triples @1080p, some settings turned down
1070ti is 95% fps of a 1080
Vega 64 is 99% of a 1080
2060 is 100% fps of a 1080 --> okay for 75 Hz triples @1080p, some settings turned down
2070 is 116% fps of a 1080
5700XT is 119% of a 1080
1080ti is 127% of a 1080 --> good for 75 Hz triples @1080p, settings full (or 60 Hz triples @1440p)
2070 Super is 129% of a 1080
2080 is 137% of a 1080
2080 Super is 145% of a 1080
2080ti is 160% of a 1080
Going from a single monitor to triple monitors you will definitely see worse fps, in the range of 30%-50% worse. I haven't seen many fps reports for ACC triple monitors and I've yet to buy the game (come on Black Friday!), but it is known for being the poorest performer and I expect my 1080 will not cut it. AMS2 will be similar to PC2 which is similar to AC and they're all slightly better than rF2 when it comes to fps. AMS is the easy one; where the other games might be borderline at some specification, AMS will be solid.
Edit: night and rain are the killers for ACC and rF2, where you'll easily drop another 30% in fps. PC2 (and presumably AMS2) take an fps hit, too, but it's more like 10-15%.
[Worth noting that from the perspective of a 2080ti, there is little difference at each step between the 1080ti, 2070 Super, 2080, 2080 Super, and 2080ti. Only 21% difference overall, but 21% of a 2080ti is 33% of a 1080!]