I'm OK with leaving it like this. It's very marginal as an entry.

Yea, I guess we could technically use the [cop] tag .

Re: Wording
It wasn't really the use of "copula" for me, I just thought it sounded odd the way it was previously. But having the tag works too in my eyes

I was going to propose changing the xref on sense 1 to refer to 如く instead, but I remembered that it was a bit special for 如し.
Nikkoku confirms that 連体形、連用形 and 終止形 were all used as 連用修飾語

I don't think there's much more left to add, is there?

@@ -22 +22 @@
-<s_inf>at sentence end; acts as copula by itself</s_inf>
+<s_inf>at sentence end</s_inf>

So, I changed the tag over [1] from [poet] to [arch] in the actual edit, but forgot to do so for the [1] in my suggestion under my comment.
I confess, your excellence; I copy and pasted the [1] in my suggestion from the pre-edit entry, take me away.

Incidentally, about [arch] over [poet]： since thanks to your references it seems clear beyond doubt that this was only used in ye olden times, it is probably best to leave it as [arch] since there is a group of archaic words that are still occasionally used in poetry - なり、ず、ざる、し、たり、etc. 
The wording on [2], I do not think matters much - if someone does not know what a copula is, it is moot either way; and I am not sure how likely it is that such a person will know and/or remember what exactly is a predicate, so that might not fare much better either.
Maybe just settle for the [cop] tag... and speaking of which, I have just added it since we are proceeding with aux-v.

(Regarding "corpi"; I said "ellision" instead of "ellipsis" even though I had correctly named it just two lines earlier)

@@ -13 +13 @@
-<misc>&poet;</misc>
+<misc>&arch;</misc>
@@ -19,0 +20 @@
+<pos>&cop;</pos>

Frellesvig, Bjarke, Stephen Wright Horn et al. (eds.) 2023. Oxford-NINJAL Corpus of Old Japanese. 
Available at: http://oncoj.ninjal.ac.jp/

In the ONCOJ, ごと is classified as follows:
"goto        (adjective; non-inflecting)        007004a (frequency 90)  be like"

These are all hits for this ごと the ONCOJ  (If you are required to enter a password, see this page: https://oncoj.ninjal.ac.jp/index.html)
https://oncoj.orinst.ox.ac.uk/cgi-bin/oncoj_tgrep2.sh?search=/007004a/

ーーーーーーーーーーーーーーーー
角川古語大辞典 tags 名詞, but that either does not account for sentence final usage, or supposes copula ellision. 
Note that none of their examples have ごと in sentence-final position.

旺文社古語辞典 第五版 performs the split you have proposed, but sadly does not give a PoS:
  １．（連用修飾語となって）．．．ように 
  ２．（述語となって）．．．ようだ

大修館 全訳古語辞典 gives 助動詞, and has only one sense:
  １．〔助動〕比況（たとえ）～のとうに。～のようだ。

Weblio古語辞典:
  １． …のように。…のよう。▽連用形「ごとく」と同じ用法。
  ２． …のようだ。▽終止形「ごとし」と同じ用法。

Nikkoku: 
（助動詞「ごとし」の語幹。本来、「同じ」の意を表わす「こと」の濁音化したもので、体言的性格をもつ）ごとく。ように。同じく。
Does not mention copular usage at all in the entry. But funnily enough they do in a note in the ごとし entry: (2)語幹「ごと」は連用修飾格や述語格などに用いられた。

As expected, the modern corpora I have access to return 0 results for this ごと.

＞ Anyway, if this is copula ellision then I am in full agreement with Sombrero and this thing is functionally a 形式名詞; but much like よう, it is not exclusively adverbial.
   Otherwise, if it never took なり then it is functionally a copular itself and the [aux-v] tag is arguably justified, or at least Daijirin and Wiktionary think so.
  ＞ From what I could gather, there existed (as a lexicalized form) only ごとくなり (cf. Haruo Shirane - Classical Japanese - A Grammar, p. 136), but not ごとなり. 
     Furthermore, not a single one of the examles in the ONCOJ show ごと being followed by なり, or any copular for that matter, so I believe it is rather unlikely to be copula ellision.
     Hence you are most likely right in your second argument, and I just went a full circle. 
     I fear Frellesvig's classification of "adjective" doesn't help us too much, since he rarely uses 国語 terminology and classification when it comes to grammar. 
     As that "kind" of adjective really has no corresponding PoS marker in 国語文法 or JMdict for that matter.

＞ Also retagging from [poet] to [arch] - while I myself have heard 如し and 如く in poetry several times, when it is ごと all I think is "maybe once or twice？" 
   but even then I might be thinking of some snippet from Old Japanese; 
   I tried looking for an example for a while but came up with nothing, feel free to change this back if you find or know of one.
   ＞ I assume some references tag [poet] because most usage examples (at least in Old Japanese) are from the 万葉集. 
      Though there are also examples from the heian and early kamakura period in references.
      Nevertheless, it is of course archaic as well, my non-古語 references are split between [poet] and [arch] (of those who give a tag).
      I do not know of any poems using ごと, as I have not really gotten to the point where I delve into such. Though I certainly wish to do so one day.


Interestingly it seems like 連用修飾語 usage was (somewhat/far) more common than 述語 usage, already in Old Japanese (only 7/90 usages in the ONCOJ are sentence final).
The majority of my (non-kogo) references describe and attest only the "のように" meaning.
Another notable, but not relevant bit of information is the fact that ごとし was far less common (according to ONCOJ) in Old Japanese, than ごと by itself (52 - 90).
And according to nikkoku, ごと was already present in the kojiki, while ごとし was not. I like to imagine that it went こと ＞ ごと ＞ ごとし, but that is just unfounded speculation on my part.


As for the note on sense 2, I find "forms copular sentence" to be a bit difficult to grasp. Perhaps what I proposed works better? Or maybe "forms predicate by itself"?
I applied a slightly modified version, with the changed PoS tags, of your proposed version. Please check whether it is fine this way.



(In writing this I noticed that I pluralized "corpus" wrong up until now...)

@@ -11,2 +11 @@
-<pos>&n;</pos>
-<pos>&n-suf;</pos>
+<pos>&aux-v;</pos>
@@ -13,0 +13,7 @@
+<misc>&poet;</misc>
+<s_inf>all senses: after が, の, or the attributive form of a verb</s_inf>
+<gloss>like...</gloss>
+<gloss>similar to...</gloss>
+</sense>
+<sense>
+<pos>&aux-v;</pos>
@@ -15,3 +21,3 @@
-<s_inf>after が, の, or the attributive form of a verb</s_inf>
-<gloss>like</gloss>
-<gloss>similar to</gloss>
+<s_inf>at sentence end; acts as copula by itself</s_inf>
+<gloss>is like...</gloss>
+<gloss>is similar to...</gloss>

The 助動詞 tag given by Daijirin is probably because of this: 「② …のようだ。「ごとし」の終止形にあたる。『逢 (あ) ふことは玉の緒ばかり名の立つは吉野の川のたぎつ瀬の―』〈古今・恋三〉」
Ironically, there is no such sense in Daijirin and I pulled this from Daijisen.
Meikyou: 「〚文語助動詞「ごとし」の語幹〛…のようだ、…のように、の意を表す。」
Weblio: 「②　のようだ。▽終止形「ごとし」と同じ用法。」.
Japanese Wiktionary has its entry for 如 as a 助動詞 with the paradigm of ○｜ごと｜ごと｜○｜○｜○ 
https://ja.wiktionary.org/wiki/ごと#助動詞

Well, the entry is missing the copulaless copular usage mentioned in some dictionaries.
That said, the dropping of the copula can happen with just about any nominal and was rather common in Old Japanese compared to modern day - you could go on for noticeably longer without seeing なり than you can without seeing だ.
Perhaps the dictionaries that say nothing of that usage did not consider it to be a property of 如 but rather just your average copula ellipsis (but do not take my word for it on this last statement, confirming or denying this would require scouring for instances of ごと followed by なり in Old Japanese - or lack thereof - which I have not done).

Anyway, if this is copula ellision then I am in full agreement with Sombrero and this thing is functionally a 形式名詞; but much like よう, it is not exclusively adverbial.
Otherwise, if it never took なり then it is functionally a copular itself and the [aux-v] tag is arguably justified, or at least Daijirin and Wiktionary think so.

I leave this rough suggestion below in case we decide to include the copular sense.
[1][n,n-suf][poet]
  [note="all senses: after が, の, or the attributive form of a verb"]
  like...; similar to...
  [see=1466930・如し[1]]
[2][n,n-suf][arch]
  [note="at sentence end; forms copular sentence"]
 is like...; is similar to...

Also retagging from [poet] to [arch] - while I myself have heard 如し and 如く in poetry several times, when it is ごと all I think is "maybe once or twice？" but even then I might be thinking of some snippet from Old Japanese; I tried looking for an example for a while but came up with nothing, feel free to change this back if you find or know of one.
Lastly, reminder to alter the part-of-speech tags accordingly, depending on whether we decide to go with it being a nominal or an auxiliary verb.

@@ -14 +14 @@
-<misc>&poet;</misc>
+<misc>&arch;</misc>