A few bugs in the WT_STRING_MATCH macro were discovered upon looking into a Coverity issue (see WT-11057 for more details).

Two possible fixes were suggested in WT-11057 - each with different semantics on how we want WT_STRING_MATCH to behave (i.e. whether we want empty strings to be considered a match or not). A decision should be made on this and documented in a comment in the code to ensure the usage and expectations of the macro are made very clear.

As part of this ticket, we should also revisit places where we may have previously been working around the potential problems with WT_STRING_MATCH (see the first comment of WT-11057 for one such example).

Finally, we should run a full WiredTiger patch as well as the required tasks in a MongoDB patch to ensure that there is no unexpected fallout.