[benchmark] Rename InsertCharacter to String.insert

[palimondo]

Applying new Benchmark Naming Convention #20334

[palimondo]

@eeckstein I'll try to kill two birds with one stone here 😜:

This can be used to validate working of #20807, because the new naming convention hasn't landed yet and the Benchmark Check will complain that the new names don't match the UpperCamelCase convention…

…but given these are quite recent additions, could we rename them for real, without disturbing the long-term performance tracking?

[eeckstein]

could we rename them for real, without disturbing the long-term performance tracking?

Renaming them will make these benchmarks useless until we do a rebase.
Therefore I prefer to apply the new naming conventions only on new benchmarks.

[palimondo]

What is the rebase in this case?

[palimondo]

@eeckstein ^^

...also, analyzing weird memory behavior of these benchmarks with @milseman in #20807 (comment), we've discovered these benchmarks need to be corrected to perform linear work regardless of N/--num-iters. If that fix changes reported times it might make sense to couple that with this rename, right?

[milseman]

Along the lines of what Erik said, a rename alone doesn't give us value in converging or improving performance for Swift. But, if a benchmark is flawed, and we fix that flaw, we should update to follow the new naming convention. This could be fixing the benchmark to be linear in time/space with iteration count, readjusting the multiplier to run in a targeted time frame, etc.

[CodaFi]

There hasn't been movement here in a year. It seems like the consensus is that we shouldn't take this without some more work. I'm going to close this out, but if you would like to pursue this please reopen it and respond to the feedback, or give us a new pull request.

Thanks @palimondo!