@@ -23,2 +23,2 @@
-<gloss>since (last month, etc.)</gloss>
-<gloss>for (20 years, etc.)</gloss>
+<gloss>since (e.g. last month)</gloss>
+<gloss>for (e.g. 20 years)</gloss>

OK, I yield.

@@ -26,6 +25,0 @@
-<sense>
-<pos>&pref;</pos>
-<xref type="see" seq="2842627">来伯</xref>
-<gloss>coming to (country, city)</gloss>
-<gloss>arriving in</gloss>
-</sense>

> it's a pretty useless exception unless we allow other kanwa-type entries that would normally combine with it, such as 来-.
I don't agree with that. Single-kanji country abbreviations often appear on their own (e.g. in headlines) so it's useful to have them searchable.
Presumably that's why several of them are in the kokugos as regular (non-kanwa) entries. But whereas the kokugos only have a select few, we've decided to include *all* of the commonly used ones. In that sense, we're not really making an exception; we're just being consistent where the kokugos are not.

Right, but Jim--who agreed with you on the other 伯 thread--is the one who suggested that this be added.  So it seems wrong to suggest that there is an agreed upon current policy.

If 伯 (etc.) is actually a kanwa-type entry and we are "making an exception for it", it's a pretty useless exception unless we allow other kanwa-type entries that would normally combine with it, such as 来-.

This has already been discussed on at least two entries so I don't really want to repeat myself, but we're making an exception for country names. That appears to be the current policy. See 2248580 and 2836257 for the reasoning.
We haven't decided to turn jmdict into a kanwa dictionary. That would be a *major* policy change.