Even if Magnepans had a perfectly flat response(and to be clear, we don't really know how the larger ones perform, and the interpretation of anechoic measurements for dipoles is not 100% clear), they'd still sound pretty different than the studio monitors that are typically used for mixing music.
But then, so do the wide dispersion Revel floorstanders. It is pretty clear in the research that people prefer things that aren't necessarily used in the mixing process. For example, (tonally accurate) early reflections, which are present to much lesser degree in a typical studio, are preferred.
Unless you are treating your listening room as if it were a studio and using main monitors that have narrow-ish dispersion you're probably not hearing the same thing that was heard during the mixing/mastering process. And I mean, there are some people out there who do try to achieve that.
But the research indicates it's not the general preference and that pretty much everyone is seeking some kind of euphonic effect that increases the spaciousness of a typical stereo recording. Really all this shows you is that stereo is an inherently busted format that needs to be "fixed" at the user's end because there just isn't enough spatial information in it to be satisfying. I think upmixing is a much better way of fixing it than fiddling around with speaker types that cannot be easily changed, but it's just an alternate solution to the same problem.
Magnepans are one way of adding some spatial effect, the smaller ones at least I don't think are a good way because they have so many issues like their problematic back wave that causes comb filtering and a major lack of SPL capability in the mid/upper bass region. But I don't see how you can characterize the overall goal as fundamentally different from a wide-dispersion floorstander. It's the same type of thing just a different flavor.
