That's the kind of workaround I had in mind and is probably sufficient, but yes, ~25 lines of cryptic and error-prone interleaved GNU make and shell script just to express multiple outputs correctly is way above my personal threshold for “too complex” or even “sane”.

That doesn't really look like "handling it" by any sensible definition. That looks like a huge kludgy workaround, almost like user implementing a build system inside the build system. (I mean, wtf, asking the user to implement a lock-and-set operation inside their build?)

GNU Make works intuitively in pattern rules, but differently for normal rules? Uh oh.

It even has builtin implicit rules. Enabled by default. Checking for the existence of files that are not found in any new project for maybe ten or twenty years (such as RCS and SCSS version control files).

On. Every. Single. Run.

Yup.

But I love the built-in implicit rules. I use these frequently.

Often, I write a little test program in C or C++. I name the file "foo.c". Then, to build it I simply type `make foo`.

Sure the rules may be a bit stale, but there isn't much software i use today that I could use in a near identical manner 40 years ago. Make has truly stood the test of time.