I think Shivraj will produce circular micro-holes.
The problem of very thin, long electrodes is that they perform a fluttering motion. This source of error can be reduced through two effects substantially. First, by the rotation of the tool electrode there will be a certain degree of stability, so that a fixed reference axis can be defined. Secondly a narrower (smaller diameter) the tool electrode is guided to a radius (eccentrically), so that an improved flushing is possible.
Of course, you will then know whether the result is sufficient for my opinion come only optical systems for application.
If you want to measure the circularity at top and bottom of the hole opening, you can use digital profile projector attached on tool makers microscope.
or you can use stacked workpieces, one upon other, clamp it tightly and perform EDM micro drilling and measure the top/bottom profiles by unstacking each sample.