Bad news: I just found an incredibly stupid coding mistake that has caused me to doubt a super important model I've been using for planning telescope pointings.
Good news: the super important model is indeed correct, and all that doubt was for nothing!