Although I agree that isocratic separations "should" produce flat baselines, that does not always happen. I would be very careful with this technique, as a very small shift in the baseline will not be corrected, and your peak size will be in error. If these are very small peaks, then this is a dangerous option, unless you look at the baseline very closely. Try to learn more about your integration settings, and maybe you don't need this option.
As for GMP compliance, from my limited reading, there is really nothing in GMP at this level of detail. Others may want to comment in more detail, but my understanding of GMP is: "You must set up reasonable rules and you must follow them!" Most auditors know little, if any, chromatography, and would be unable to tell if you were doing things correctly or not. They are more concerned about whether everything is documented and signed.