I don't have any objections to the edit, but re: the [rK] tag at 4%, I just assumed that most of the n-gram counts were from personal names. I should have included that explanation, though.

gg5
https://uk.inaturalist.org/taxa/461556-Osmunda-japonica
https://en.wikipedia.org/wiki/Osmunda_japonica
ぜんまいが咲いた	0
ゼンマイが咲いた	0
薇が咲いた	53
---
ぜんまいを采る	0
ゼンマイを采る	0
薇を采る	        33
---
ぜんまいの煮物	2916	83.0%
薇の煮物	        88	2.5%

ぜんまい is also 発条 (spring) so 薇 is more common than the raw counts suggest (but I wouldn't have tagged it as rK even at 4% of total counts).
The kana forms still dominate in culinary contexts. 
I can't find a source for "osmund". "fiddlehead fern" doesn't refer to a species.

@@ -6 +5,0 @@
-<ke_inf>&rK;</ke_inf>
@@ -22 +21,2 @@
-<gloss>royal fern (Osmunda japonica)</gloss>
+<gloss>Japanese royal fern (Osmunda japonica)</gloss>
+<gloss>Asian royal fern</gloss>
@@ -24,2 +23,0 @@
-<gloss>osmund</gloss>
-<gloss>fiddlehead fern</gloss>

Google N-gram Corpus Counts
╭─ーーーー─┬─────────┬───────╮
│ 薇　　　 │  28,728 │  4.1% │
│ 紫萁　　 │       0 │  0.0% │ <- daijr/s, meikyo; adding
│ ぜんまい │ 507,199 │ 71.8% │
│ ゼンマイ │ 170,209 │ 24.1% │
╰─ーーーー─┴─────────┴───────╯

@@ -5,0 +6,5 @@
+<ke_inf>&rK;</ke_inf>
+</k_ele>
+<k_ele>
+<keb>紫萁</keb>
+<ke_inf>&rK;</ke_inf>