It's been good to see more and more linguists recently pointing out how deeply embedded modality chauvinism still is in our field, where language is often used to mean 'spoken language' as if visual and tactile languages don't exist. My contribution :

I recently read this new paper on languages on islands. It's interesting, but it's about spoken languages only. One of the main observations is that there are very few language isolates on islands. Not true of signed languages! https://www.researchgate.net/publication/340244404_On_languages_on_islands

Just in the Caribbean, I know of at least five sign languages which were created on islands without contact from outside (there are almost certainly more), and there are other fairly famous examples around the world (Kata Kolok, Martha's Vineyard SL, Miyakubu SL).

I noticed that the authors of the paper used @Glottolog as a reference point for which languages are isolates. Because all sign languages there are listed under the 'Sign Languages' 'pseudo-family,' none seem to be identified as isolates, though many surely are.

I don't think tactile languages are represented at all in the classification. I mean, I can see ways in which it might be challenging to decide how they should be classified, but surely some should be represented somewhere in "a comprehensive catalogue of the world's languages."

As @linasigns, @AnneliesKusters and others have pointed out, the typologies which have been proposed for signed languages need much more work. Historical relations among signed languages are not well understood, global sign language diversity is largely not understood.

Some signed languages may not fit well into the genealogical type classifications (nor many contact languages).

Anyway, surely time for a bit less hubris about how comprehensive any current typologies of human languages could be, and definitely time to start saying "spoken language" when that's what you mean.