I see a few issues with a system like this, as opposed to LOG/DEBUG/INFO/WARN:
 too few levels (this is only a problem on bigger systems)
 it needs to be explained (one could probably figure most of it out given enough of the logs, but not based on limited chunks)
[*] non-standard: doesn’t work with existing frameworks, isn’t immediately familiar, and won’t work with tools that expect logs formatted in another way
Nonetheless, the succinctness and simplicity are appealing enough that I might start doing this for my own projects. This is an awesome system for its intended use.
I guess I should have clarified – this is for software I “own”, i.e. mostly my own personal projects and internal tools.
It may not be immediately familiar, but now I have an explanation online I can point people to ;)
It is pretty clear that you only use it internally; I was just trying to hypothesize whether it would be feasible to use this scheme on a wider scale, and I think it would if you could gather a critical mass of code that follows it.