OK, I'll split it off.
Re "the other entries need to be split, too.", I'm not sure I agree. They are not such extreme outliers.

@@ -10,6 +9,0 @@
-<k_ele>
-<keb>子生婦</keb>
-</k_ele>
-<k_ele>
-<keb>幸運夫</keb>
-</k_ele>
@@ -22,0 +17,3 @@
+<reb>こぶ</reb>
+</r_ele>
+<r_ele>
@@ -25,3 +21,0 @@
-</r_ele>
-<r_ele>
-<reb>こぶ</reb>

I'm neutral on it, but if this is done, the other entries need to be split, too.

Daijs: [補説]「よろこぶ」に通じることから祝い事に用いられ、その場合は「子生婦」「幸運夫」などとも当てて書く。
昆布	2150680
恨布	< 20
子生婦	1417
幸運夫	< 20

(a) I think 恨布 can go. 0 n-grams and 56 Googits, most of which are from here.
(b) 子生婦 are 幸運夫 obscure outliers only relevant to betrothal gifts. I think they'd be better off in their own entry with an appropriate gloss. They are not used for the regular food ingredient.

@@ -9,4 +8,0 @@
-</k_ele>
-<k_ele>
-<keb>恨布</keb>
-<ke_inf>&iK;</ke_inf>

daijs https://kotobank.jp/word/昆布-503957
https://www.tofugu.com/japan/proposing-to-a-japanese-girl/

"In modern Japanese, ateji principally refer to kanji used to phonetically represent native or borrowed words with less regard to the underlying meaning of the characters. This is similar to man'yōgana in Old Japanese. Conversely ateji also refers to kanji used semantically without regard to the readings."

Not really ateji as the meaning of the characters is important in these betrothal gift words. It's really an instance of phono-semantic matching.
https://en.wikipedia.org/wiki/Phono-semantic_matching

@@ -12,0 +13,6 @@
+</k_ele>
+<k_ele>
+<keb>子生婦</keb>
+</k_ele>
+<k_ele>
+<keb>幸運夫</keb>