Why is this editorial?

I don't see any alternative reading that makes sense.

I see three potential readings of the status quo:

1. The sentence constrains the widths of only the types listed in the table. This is what I think is the intended reading.
2. The sentence constrains the widths of all integer types, including extended integer types. This is the reading taken at [conv.integral] Conversion of bool to width-1 signed integer type CWG#365 (comment). I don't think this is intended (if it was, it would be explicitly stated, or there would be another entry in the table reading "any other signed integer type" or similar). If this reading applies, it's not clear which minimum is meant to apply to an extended integer type (the linked comment assumes 8 but it could just as well be 64).
3. The sentence prohibits the existence of extended integer types. This is the reading suggested at [conv.integral] Conversion of bool to width-1 signed integer type CWG#365 (comment). This is obviously not intended.

Maybe I'm wrong and 2 is what is intended? Or maybe it's not clear enough for this to be editorial, in which case I'll happily file an issue.

@jensmaurer Thoughts?

The table lists widths of the particular standard signed integer types. It's editorial to clarify the reference to that table (and thus, to this particular list of integer types) to be the "standard signed integer types", because that's what they are per the definition in basic.fundamental p1.

Sounds good, thanks!

@tkoeppe, let's go for it.