Cardioid speakers in real (home) rooms

[illusonic]

Yes, from Sd perspective the chances that O11 will get louder are quite obvious, the small LS50M isn't made for that.

On the other hand directivity is not just the spinorama which averages all angles (and you cherrypicked by using the 30° one as the 0° looks quite worse) but the whole plots matter and should be viewed, here in comparison:

View attachment 546626 View attachment 546625

View attachment 546627 View attachment 546628

View attachment 546629 View attachment 546630

It is obvious which one is clearly smoother.

Normalize these curves to 30° and the result will be much better. With the speakers pointing straight into the room, 0° sound will almost never reach the MLP.

 

[thewas]

Normalize these curves to 30° and the result will be much better. With the speakers pointing straight into the room, 0° sound will almost never reach the MLP.

The lower plots I have shown are normalised (to 0°) and always problematic angles will reach the listener in form of reflections.

 

[illusonic]

The lower plots I have shown are normalised (to 0°) and always problematic angles will reach the listener in form of reflections.

Yes normalized to 0° instead of 30°. It's a matter of the reference axis. And I did not cherry-pick that axis. As I explained, 30° gives better image width. So it's the better axis overall.
What I tried to say with "0° sound will almost never reach the MLP" is that due to 0° sound mainly bouncing between front/rear wall it takes many reflections till sound may reach the main listening position. By this time influence on tonality is negligible. Now compare this to floor/ceiling bounce of popular non-coaxial monitors. Much bigger effect on tonality.