I don't have specific experience with this method, so someone else may have to fill you in better. Offhand, I wouldn't deviate in such a manner, on general principals. Its very plausible too much analyte will saturate the reaction, and misrepresent the results. It would even give me pause if you kept the sample, but altered optical parameters (narrowed slit, shortened scan, or reduced detector sensitivity)
You could Google for the technique, and see if you find a reference that says just what you propose, but I'd doubt someone would submit for peer-review something like that. You could get lucky 'tho, just check them all and see if someone has deviated in just such a way.
Of course, you can always test it for yourself -- prepare a bunch of standards at double strength, diluting before and after assay, and see if the results are comparable to your satisfaction.
This really is all up to the rigor your application demands.