@snakeeyes
So the reference I was thinking about is Dickason's LDC v7. Specifically what he discusses is a comparison of WTW (d'Appolito) as a 2-way vs the same configuration with a 2.5-way with one of the woofers (lower) crossed 1 octave lower than the other which crosses normally to the tweeter. In this instance, his modeling showed that the lobing you would experience in a 2-way WTW was lessened with the 2.5-way at the cost of increased asymmetry in the vertical dispersion. His contention is that the lobing in a traditional 2-way WTW is the lesser evil and is less audible perhaps than the resulting asymmetry in dispersion from the 2.5-way version.
His modeled polar plot for vertical dispersion showed a decided upward tilt in the 2.5-way speaker. His subjective listening experience using the same speaker, one set as a 2-way, the other as a 2.5-way, played mono and compared side by side, was that they shared a similar timbre, but that the image depth was lacking in the 2.5-way.
I cannot say whether this effect might carry over in any way to a speaker like the F35 or F36. I don't have the experience building or testing speakers to know what to expect. I think it is interesting though that Dickason dedicated a small space in his book to discussing this in terms of why and where drivers are placed on the baffle and the resulting effects.
My instinct is that using matching drivers with different XO points is not optimal, rather a comprimise to tease performance out of a complex system. In my mind, if you are going to the effort of building XOs with 2 different high-pass filters, you might as well go to the extent of putting a mid-woof in and building a bandpass filter, thus taking the risk of crossing woofers too high or tweeters too low completely off the table.
Perhaps
@shadyJ can shed some light on this?
Not so much geared toward a specific speaker, but in general terms?
Cheers!