User blog comment:B1mb0w/Growth Rate of the S Function/@comment-5529393-20160712123611/@comment-5529393-20160712132508

That's pretty tricky; it never occurred to me to use that substitution rule, since it certainly appears that "restricted" and "generalized" apply to the form of the S function, and not just it's value. So I naturally thought that a "generalized S function" would have to have the form S(a,b,c) with the condtions you state.

The goal of a description should be to make things as clear as possible to the readers, so you should avoid doing tricky things that will fly over your reader's heads.