Modelling two-way bookshelf with woofer on top

I'm looking at modelling some two-way speakers that will be placed on a 24" wide shelf next to a TV.

Normally a two way driver is modelled with the microphone in line with the tweeter and then the woofers' acoustic Z offset is determined.

At one stage Mission used to have speakers that had the tweeters at the bottom which to me seems to be logical: A mid/woofer will start to beam at higher frequencies while a tweeter has wider dispersion. As such it seems logic to me to want to listen at the woofers' height.

It also seem to me, but might be mistaken, that if the speaker is to be modelled with the woofer on top and the microphone in line with the woofer that then less Z offset required and hence results in a different crossover.

Am I correct in this assesment and are there disadvantages to having the woofer on top that I overlook?

Thank you for your input.
 
This is what I do on my transient perfect 10F/RS225 speaker. The mic axis on woofer gives the tweeter added delay. Sort of like a stepped baffle without the step. It works well for a FAST/WAW type topology with the woofer going up to cover the mid bass and mids. I cross at 900Hz. In my case, the offset is good for circa -3in equivalent setback. With a 1st order crossover, the distance is not so critical and the time alignment is good for a wide range of listening positions.

https://www.diyaudio.com/community/threads/10f-8424-rs225-8-fast-waw-ref-monitor.273524/

1667998417150.jpeg


This is the sort of step response I can get over a wide range of positions.
1667998524058.png