In the first case, the results would not have been biased because spilling the sample does not effect the analyte to internal standard ratio. Furthermore, adding more MIBK also does not effect the analyte to internal standard ratio.
In the second case, the results would be biased because the internal standard had not yet been added.
A good discussion of the internal standard method can be found in "Modern Practice of Gas Chromatography" by Robert Grob, Ed.