Certainly, it could be done. I first used an internal standard to remove systemic errors from an HPLC analysis I was doing. But even in that case, the real source of the error was my sloppy dilutions -- I was new to pipettors in those days. So yes, if you suspect there are errors in the dilutions, or handling, or if the instrument autosamples, or has a systemic optical flaw, yes you can do it. Don't know if that sort of thing comes up of often though.