In THEORY this would work...unfortunately, the real world isn't as kind.
I think it depends on a lot of factors:
Firstly, if you analyse the remaining reactants instead of the products directly, you have to be EXTREMELY careful about how much you put in to begin with. Depending on the GC apparatus you're using...it could be as sensitive as picograms - which is a VERY difficult sensitivity of get measured at the beginning.
Also, depending on what kind of GC apparatus you have (you may have just a GC-FID, maybe a GC-MS, GC-TOF-MS, GCxGC-TOFMS) then that should factor in your decision as well.
Final comment: I'm not familiar with calibration of GCs...if it was me, I'd be running a few GC runs with Standard additions