Anybody else's spreadsheet that's complex enough that you couldn't build your own in an hour or less is probably full of unstated assumptions and other traps.
The spreadsheet could deal with various sets on inputs; say there are three things with three options each; call them a,b,c; d,e,f; and g,h,i. While all nine can be used in various combinations, the spreadsheet may only get 23 or 24 of the combinations right. The designer having decided that the rest are non-credible combinations, and them being so non-credible that there's no error trapping for them. But you start using it and you think that one of those heretofore non-credible combinations is a nifty solution to something. The results may be other than what's desired. You may have gotten so many good results for combinations that were anticipated that you entirely miss the mess-up made when you picked that combination.
And you own the results. The spreadsheet author has no liability, it's all on you. Far better to write your own. Debug it yourself. Understand what assumptions went into it. Know what cases produce rock solid results vs. what cases might need additional verification vs. what cases should just be completely avoided.
I’ll see your silver lining and raise you two black clouds. - Protection Operations