> I am still not sure that it is worth to add 50 lines of the C code to optimize 0.7% of calls.

As I stated, the reason that this is 'relatively' lower is that there are just too many calls, so it just allocates a rather small part of the total pie. 

> Does it include tests?

Yes, to some extent (some packages include tests, some don't). 

> function calls with 3 or 4 constant arguments. Does it include tests? Could you please show several (or several hundreds) non-test examples?

Here are 2 gists (attached together) that contain the locations where such function calls are performed within a loop (the first one contains your first query, 3 and 4 arguments and the second one contains 5+). 

I added an extra guard to check whether filename starts with test_ or not, so it should be generally free of tests (at least the ones that follow the general convention). I didn't deal with doing an extreme analysis, it just checks whether there call is made under any for/while's loop so there might be some false positives. 

https://gist.github.com/isidentical/8743aca3f7815cd19ee71579ca5ba974