E-M5II extreme macro for metrology

Interesting question. Disclaimer: since this is an engineering question, you get an engineering type of answer that may not be a perfect match with a photography audience.

About diffraction: when modeling the sensor resolving power in combination with diffraction limits, the 40 Mp MFT camera will suffer a 20% impact on resolution because of diffraction around f/6. If you do not care about depth of field (flat subject), I would advise to stay at f/4 or below for MFT. This advice holds equally for the D810 to achieve the same resolution at subject plane (at magnifications larger than 1:1, the diffraction limit is at the subject side). Take also into account that the real aperture at maximum magnification is likely one or 2 stops worse than the value specified for focus at infinity. So basically, shoot wide open. If you get more resolution by stopping down, there is still significant room for resolution improvement by finding a better lens.

About lens data: I have only tried the Panasonic Leica 45/2.8. At maximum magnification (1:1) the lens is actually sharpest (in the center) wide open, which suggests very low aberration levels. At this aperture I can, on a GH4, clearly resolve a tiny "plus" sign with an overall width of 20 micron and a line width of 4 micron (chrome on glass). I do not know the performance of the Oly. Take chart measurements for comparing it to the PL45 taken from several meters with a grain of salt. They are not representative for macro lenses since they can be optimized for closer distances.

I don't know what type of structures you need to resolve, but my model of the GH4 reaches for a perfect lens an MTF50 value for repeating line/space structures at 4 micron line width. This goes down to 2.5 micron when i substitute in the same model a 40 Mp sensor with similar anti aliasing filter strength (additional blurring as a percentage of pixel size). Based on the model in combination with the measurement, I would take the bet that the PL45 in combination with the EM5ii will resolve better than 4 micron line/spaces when placed on a very sturdy tripod, but I'm not sure whether you will reach your quoted 2.