I enjoyed this article http://eli.thegreenplace.net/2017/restructuredtext-vs-markdo... and since the article didn't provide supporting evidence for the Linux/OpenCV/LLVM assertion: https://www.kernel.org/doc/html/latest/doc-guide/sphinx.html and http://docs.opencv.org/2.4/doc/tutorials/introduction/how_to... and https://github.com/llvm-mirror/llvm/blob/master/docs/index.r... respectively

I do think rST is waaay more expressive, but I also recognize that in many of the instances one would want to use markup in a chat or PR situation, the expressiveness likely wouldn't be well received if the trade-off is verbosity.

This is something in life that I file away with competing regex standards: my brain just has to switch languages based on the app in which I'm typing (between markdown, pseudo-markdown (ahem, Slack), org-mode, rST, etc).

For personal documentation (assuming you mean notes): follow your heart. Personally, though, I wouldn't use RST for any non-python public documentation at this point, but for my personal notes, it's hard to beat the extensibility of RST.