FIBIGRUS Yes, I do think Pianoteq does sympathetic resonances a bit better. It seems to me that modeling has an edge for things like this since resonance emerges from the interaction between strings and it is impossible to record every combination of all strings. In Modern D they start with samples for the resonance but they are probably also using modeling-like methods to get to what you hear in the end.
The resonance implementation in 1.1 is good enough for me. I like the base sound of Modern D better so I'm ok with it. You can find info in this forum on sidechaining PianoTeq to get the resonances, I think @OlivierW knows more.