康熙字典（內府本）has 𩹆, which is reasonable since 孱/孨 is a pair of semantic variants.

https://nrs.harvard.edu/urn-3:FHCL:4339956?n=3540

What is the version of 康熙字典 in this evidence?

I agree with Huang Junliang that full edition information should be provided, and it is not sufficient to simply give the source as "康熙字典". The 中华书局 edition p. 1478 has 𩹆, so the proposed character is certainly an error form. I do not think that we should encode error forms of characters from poor-quality editions of the Kangxi dictionary.

▲ 《大南一統志》，卷之四

This character is also found in the ancient Vietnamese book, which reads as chuồn.

We will propose a horizontal extension for the Vietnamese character after this has been encoded.

The new evidence supports encoding as a V-source character, but I still think the original China evidence is an error, so there is still insufficient evidence for encoding as a G-source character.