The next evolved-over-multiple years large and complex spreadsheet tool that I inherit that has even a rudimentary test suite will be the first.

More often than not, the fact that a spreadsheet was used meant business was doing an end-run around lack of IT support initially to build the thing and wanted the quickest thing that appeared to work without any concern for (or even knowledge of) software development process, and its almost certain that in the history of changes (even if IT was forced later to adopt the spreadsheet) have been do the minimum required to meet each new requirement without really understanding what went before (often because the one person that understood what the original was doing has left.)

Which isn't really a problem with the tool, but with the social conditions which make the tool attractive to use in the first place.

"good spreadsheet"

I think the biggest problem is the small crossection of these two terms.

there is a cross section who knew :-)

In the past a company y that I worked for went down and one of the factors was some wacky spreadsheet in our finance workflow :-(