I don’t understand why mixing and monitoring can’t be done on an identical system although preferably not by the same person.
Here are some of my experiences:
Isolation creates more low frequency issues. Unless you have performers (or neighbors) in the next room, I’d avoid it!
Diffused, flat frequency response reflections off the walls makes a big improvement. We learned this at Motown from RCA in the mid 1960s. I currently have a 4 inch thick foam absorber living behind my head because I need to be close to the wall. My only other room treatment is my collection of about 2000 LPs.
Felt diffraction treatment around drivers can make speakers both measure and sound flat without needing magic high frequency correction curves. I also avoid ported speakers like the plague. They interact with a room at low frequencies far more than speakers in sealed cabinets do. My first experience of this was a pair of BBC LS3/5as in 1976 and my current Duntech Sovereigns work exactly the same way only with LF response down to 27 Hz..
This was also something we learned about at Motown in the mid ’60s. We called B&K to ask why eqing our control room speakers didn’t sound right. We found out that nobody they knew of had ever tried this before. The microphone designer didn’t think it would work outside an anechoic chamber because of diffraction in both the microphone and the speaker drivers. I moved to San Francisco in 1972 and watched “name experts” screw up room eq. to the point that people were getting better mixes monitoring with KLH 6s, JBL L-100s and Yamaha NS-10s than their ten thousand dollar mains! That eq insanity ended with Dick Heyser’s Time Delay Spectrometry that solved a lot of the measurement issues but it has reared its head again in recent years.