Document .GetLeftPart doesn't do encoding for already encoded unicode characters

Hi,
Previously I filed https://github.com/dotnet/dotnet-api-docs/issues/11253. Another undocumented quirk is that the .GetLeftPart method knows if a Uri is already encoded, so calling it again doesn't result in double encoding. This should be included in the official documentation. Discovering small undocumented behaviors during runtime was not a great experience.
We were under assumption it always to encoding based on previous issue.
Overall, this is a useful method because it removes the SAS token from the URI.