Canon 70-200 f/4 L IS with 1.4x or Canon 70-300 f/4-5.6 L IS or 300 f/4 L IS over Tamron 70-300 VC?

dheerc wrote:

At 200, the 70-200 is slightly better wide open and at f/5.6, it is much better than the 70-300L and obviously better than the Tamron as someone pointed out.

For some reason he didn't test the other two at f/5.6 on crop. You can;t compare the f/5.6 scores to the wide open or f/8 scores. On aps-c f/8 already has diffraction hitting and knocking out max sharpness and starting equalize lenses more in terms of resolution (although not for CA and perhaps not even for global contrast richness and color either) so if the lens is good, the f/8 score at the center will usually be worse than at f/5.6. And most lenses are not quite the best at wide open so f/5.6 is almost always sharper too.

At 200mm wide open the difference probably come down to copy variation. With my copies the 70-300L was a tiny bit better wide open 200mm than the 70-200, with his copies it was reversed, but a minor difference in both cases. Where it gets weird is on TDP where he gets the 70-200 easily winning which doesn't seem to match to typical copies and barely any copies at all and his tamron result seems even more out there (although it sounds like you might be one of the ones to get results more like his).

From what I've seen on various reviews and blogs, the 70-300L is usually a trace better at 200m and better at 70mm than the 70-200 and worse in the mid-range although some get it reversed where their 70-200 is a bit better at the extremes and a bit worse in the center, this seems to be much rarer though it seems like a 6:1 ratio have it more like how it was with my copies. I think TDP is the only place I've seen the 70-200 do noticeably better at 200mm and do better even with the TC at 280mm.