Greg this idea was more to suppress the effects of the optic on very bright stars. No matter how good an anti reflection coating is or how well an optic is made, a very small amount from from a very bright star will be in the order of all the faint stuff we are trying to image. This is even more true with modern front illuminated very sensitive sensors.
The formula is
very small number X brightness of very bright star = measurable signal
All of experimental science is about signal to noise. In any experiment the major aim should be to minimise noise first. Not just collect data. It is pointless when the data you are after is below the noise you are collecting.
Bert