Is There a Universal Model for Predicting All Molecular Geometries?

[DeathKnight]

I read somewhere on net that hybridization, which is a good model for most of the compounds, is not entirely correct i.e. some molecules show considerable deviation in there molecular geometry from the predictions of the hybrid model and we only use it because it is easier way to understand the molecular geomerty of a great number of molecules. My question is; Is there a model which is true for 'all' the compounds? I tried google but couldn't find anything useful.
Any help will be greatly appreciated.
Abdullah

 

[FunkyDwarf]

i always thought molecular orbital theory solved those problems ie no longer need resonance for benzene. have you looked at that?