My question is quite naive since I am not into the source reconstruction.
I am just wondering the influence of a template MRI vs individual MRI on the source reconstruction for oscillatory activity (beamformer ? ) ? I cannot find any references or discussion on this topic.
Will the reconstructed activity be really different ?
When using a template, the location of the reconstructed sources might be incorrect (in a way that I can't tell you how to quantify), but you should be able to make similar observations in terms of timing (which is what matters the most in MEG/EEG).
This might be a nice reference paper for modelling errors in forward and the effects on the inverse source localization: