If X extends or composes well in a toolchain that people already use, X gets multiplicative impact to the (adopting fraction of users of the) toolchain rather than the impact you directly expect from X.
This sort of misprediction of impact is a bit of a pattern in my experience, so much so that to counteract it I ask questions about toolchains and composability when looking at things.
Comment inspired by an HN thread, which I won't link due to no shaming policy, on how a particular type of look-and-feel for Graphviz illustration is harder than just hand-sketching the same LaF, without appreciating that there exist, today, N things which export Graphviz.
(c.f. sketchviz.com/new for the actual artifact under discussion though.)
