As the paper points out, this an old problem. Excel is an excellent tool unlike most Microsoft applications. You learn to live the quirks. When you see the "September" genes, you just file it under "do a final cleanup and be sure this doesn't wind up in the final submission of the paper".