I happen to like developer tools like build systems. I know that for some, working on the build system is like eating your vegetables. This paper offers a framework to evaluate different build systems. The writers identify several judgment criteria:
Build systems vary on many axes, including: static versus dynamic dependencies; local versus cloud; deterministic versus non-deterministic build tasks; early cutoff; self-tracking build systems; and the type of persistently stored build information.
The paper is dense, and I'll have to read through it again, but my first takeaway is to evaluate Bazel.
PDF from cambridge.org