A very basic point that is almost always overlooked when discussing SEM “resolution,” is the difference between “resolution” and “resolving power.”

An SEM might have a resolving power of 1 nm or so based on the observation of gaps between gold particles on a featureless carbon film. Fine—if ones job is to look at gold on C.

For the actual 3-D samples most everyone looks at, EVERY micrograph has a DIFFERENT resolution, closer to 10 nm, because of secondaries generated by back scattered electrons, edge effects, …etc. If you want resolutions closer to the instrument's resolving power you have to look at thin TEM specimens without sharp edges and angles.

In our semiconductor lab we have several FESEMs advertised to have resolutions of 1 nm that can not resolve 4 and 5 nm layers in a film stack.