A family of Microsoft spreadsheet software with tools for analyzing, charting, and communicating data.
> simple decision trees to determine who gets what bug or suggestion
Before determining who, I'm pretty sure they determine whether (to look at the bug or suggestion). And I would be not surprised if in that determination they employ algorithms created for them by Nobel winners, which find out, eventually, something like (degree of increased satisfaction from the improvement) * (number of satisfied paying customers who will notice). In that process, the vast majority of suggestions will be thrown out.
I participated in many beta programs, for Microsoft and others, and in each beta test 99% of all submissions are not bug reports, but ideas how to improve the product, even when the developer is specifically asking for bug reports and is well past the phase of determining feature set. You can't even imagine the stream of ideas every big software maker is bombarded with every day from all directions, each submitter being confident that he knows how Windows should work better than Microsoft and all other Microsoft users combined. To find out that 1% of them that actually makes sense and that is requested by at least two users, is very nontrivial task. To further find out that 1% of 1% that you have the resources to implement, is another.