That is an interesting way of looking at it, but it doesn't bring out the key point that the pyridine lone pair has the wrong symmetry to contribute to the pi system. In the right-hand structure for pyridine, if it were possible, the double bond to the left of N must lie in the plane of the ring, and cannot contribute to the aromaticity. (Eight pi electrons would not be aromatic anyway.)
Also, the left-hand structure for pyrrole is not sp3 hybridised - pyrrole is known to be planar about the N. It is sp2 with the lone pair in a pz orbital. In fact, I think it is a general rule for drawing resonance structures, that you are not allowed to change the hybridisation between contributing structures. For the neutral structure, there may be an energy cost for changing tetrahedral sp3 N to planar sp2 N, but it is compensated for by the fact that the latter can be stabilised by resonance; hence the molecule is planar.