The trouble with math models is garbage in = garbage out.
Not trying to be difficult but unless one can proove all steps along the way - the result may be meaningless. So on models I just cut n try and try to remember each result .
The popular cheeks are a waste of time -in my book
The flying flounder look has real advantages tho - if side area distribution is well balanced
as far as fuselage adding downline drag- another waste of time .
Going electric is a BIG step forward because constant speed (which yeilds more constant trim setups) is a BIG advantage. On the models you show , the flounder would be easiest to trim but my approach is to reduce decalage to absolute mins FIRST then balance till the model gets wandering tendencies.

Then if tucking is a problem - I would mix as required
the roll couple - add dihedral till it stops -if possible- or add mix.
This all works for me but I no longer compete in Pattern - so the subjective pattern stuff is of no interest to me
I just want a model to be easy and predictable in flight
going for zero input in any atitude - may be possible (tho I think not) .