That's a very good philosophy, and one which can be largely verified via measurements, which makes it especially attractive.
Another philosophy is that "a desirable speaker conveys the perception of experiencing a live performance." This too implies good measured performance but might even go beyond targeting "what the mastering engineer heard". For instance, prioritizing the live performance experience over the mastering engineer experience leaves the door open to upmixing a stereo recording and playing it through a multi-channel system.
I think that statement is arguably applicable to mixing monitors, which are tools for detecting and correcting problems in the mix. To this end, two or more significantly different-sounding mixing monitors are often used during the process. Some high-end active monitors even have presets which mimic the characteristics of various lesser but considered-to-be-useful mixing monitors.
But I do not believe that statement applies to mastering monitors. Let me explain: One of the most important jobs, perhaps THE most important job, of the mastering monitor is to "impress the clients", with the "clients" being the musicians who are paying the big bucks for that studio to mix and master their music. When they show up and want to hear what their new album sounds like, you want them to be blown away by how incredibly good it sounds.
I think bad speaker/room interaction is commonly an "enemy", but I don't think rooms themselves are categorically "the enemy", at least not if spaciousness and envelopment are desirable attributes. My recollection is that Floyd Toole points out how good reflections provide the ear/brain system with multiple "looks", which aids in deciphering complex sounds. The phenomenon of soundstage widening due to the sidewall reflections requires sidewalls. And my observation has been that spectrally-correct in-room reflections can contribute to what might be called "timbral richness", which is one of the attributes of a good recital hall or concert hall.
I think it's possible to design a speaker to largely work "with the room" rather than "against the room", but that's not going to be the same speaker for each and every room, and definitely not for each and every set of priorities.