Let me clarify one other point, since you were asking about mass-based calculations. In general there is a cubed relationship between protein mass and physical dimensions, of course. However, there is a lot of variation based on shape. That is, if you are referring to a globular protein, then going from 100 to 800 kDa would only give ~2x larger linear dimensions. However, if the protein were a filament of fixed cross-section, then increasing the mass 8-fold will also increase the linear dimensions 8-fold.
So, generally speaking, proteins in the 100kDa-1 MDa range will be ~100 - 200 Å in size, but there can be considerable variation based on shape, so the practical reality is that you need to look at your images, identify the target of interest and measure its longest axis.