[css-values-5] Special case equal progress/start/end values in progress functions?

[cdoublev]

Per spec, the result of progress(1 from 1 to 1) is NaN, which might be censored to 0 at the top-level of a calculation tree (#10979).

This case is very unlikely, but I wonder if it should not also resolve to 0 when it is used in another calculation, rather than infecting it with NaN.

[tabatkins]

Ah yeah, this should be special-cased. It being a divide-by-zero is just an accident.