.casefold() vs. .lower()

I think the method .casefold() was not yet introduced, but I saw it when Craig introduced string methods with "help(str)." It seems like .casefold() allows caseless comparisons by more aggressively changing characters to the lowercase than .lower() does. That is, suppose you have smpl_str = "der Fluß". Then smpl_str.lower() == "der fluß", meaning that the character "ß" is still upper case. In contrast, smpl_str.casefold() == "der fluss", which has all characters in lower case.

What are the pros and cons of using .casefold() versus using .lower()? I understand that people generally use .casefold() when characters other than Latin or Greek characters are expected, but .lower() when expected characters are restricted to Latin or Greek characters.

However, I haven't found a satisfactory answer as to WHY that is. In other words, why not use .casefold() all the time since we should, as Craig says, "expect the unexpected"? Is there any reason to use .lower() rather than .casefold? E.g., is it much faster/efficient?