I've been mulling this over for a while, and I've come round to Robin's position.

@@ -15,6 +14,0 @@
-</sense>
-<sense>
-<pos>&n;</pos>
-<xref type="see" seq="2026130">西語</xref>
-<misc>&abbr;</misc>
-<gloss>Spanish (language)</gloss>

I think their use as abbreviations for languages is very limited. All that comes to mind is titles of dictionaries (e.g. 和仏辞典) and translations (e.g. "独: Wissenschaft"). Unless the name of the language differs significantly from the country name (e.g. 英 - United Kingdom/English), I don't think having a language sense adds much clarity.

I think adding language senses would make the entries clearer, as it might not be always obvious from context alone.

Most of these single-kanji abbreviations for countries can also represent the languages of those countries but only a couple of entries currently have a "... (language)" sense (or gloss).
Either we should add language senses to 日, 独, 伊, etc. or we should drop the language senses/glosses. My preference is the latter. It's almost always obvious from the context whether these abbreviations are referring to the country or the language. I note that the kokugos only mention the country.

@@ -12 +12 @@
-<xref type="see" seq="1766800">西班牙</xref>
+<xref type="see" seq="1766800">西班牙・スペイン</xref>

Koj: "西班牙スペインの略。「米西戦争」"

@@ -11,0 +12 @@
+<xref type="see" seq="1766800">西班牙</xref>
@@ -14 +15,6 @@
-<gloss>Spanish language</gloss>
+</sense>
+<sense>
+<pos>&n;</pos>
+<xref type="see" seq="2026130">西語</xref>
+<misc>&abbr;</misc>
+<gloss>Spanish (language)</gloss>