We of course routinely did that sort of metrology at work, aerospace optics. One generally predicts, and later measures the telescope (sans filters) implicit in your attached "mirror" and "optics" curves up there. One would then factor in the separately-characterized detector and various filters, etc. and also the anticipated environmentals. And, once on orbit, the system would get further cals on such things as known stars, the sun, and controlled artificial calibration targets on the moon or earth. Stuff like that is routine, if you actually want to use your instrument as an absolute "quantitative spatial radiometer" which is fancy-speak for a camera that takes really pretty pictures of the cosmos! Many imagers also have "upon demand internal / on-board" radiometric calibration sources. I'm on a couple of patents of such things... and used to ~calibrate the calibrators~.
I set up measurements on some hobby telescopes and components. Couple of those shown here >>> Tom