Ann: AGM Presentation, page-53

I saw that and thought exactly the same thing. We see on p13 what looks like a corner of a square membrane with presumably 4 corner attached and centre pivoting flexures. This is also shown with colour coded compliance (movement? acceleration? deviation?) on p12. Having multi-physics modelling would allow such things to to be modelled and evolved readily. If you look back at earlier movement tracking images they show a lot more deviation, and if I recall I read about this being a large part of the problem with the audio/acoustic quality shortcomings of the earlier design.