It is not an average, protein with change in amino acid number or type would be distinguishable. Deletion of one amino acid can not be "missed", such a protein would be seen as another peak in the spectrum. You are twisting my words, 1 to 0.001 is not % variance, but absolute one so it refers to 1 - 0.001 Da.
I would prefer some link to relevant and official literature as your answer to this question.
What other and more conclusive method than MALDI would you suggest to confirm the sample is GH? AAA is outdated 70 years old method. Still good for some applications, but not for this.
What does the AAA method tells you about the sequence? Absolutely nothing. The only information you get is relative ratio of amino acids in the protein. Moreover it might be strongly biased, because some amino acids decompose over the hydrolysis step (boiling concentrated hydrochloric acid for about 20 hours). To find out the exact sequence you would need to use protein sequencing or at least peptide mapping.
Actually there were two methods used to confirm the sample, is a protein.