This comment is from individual expert Ma Shijie：

T1092_.20.0276a12
（唐） 菩提流志 譯《不空羂索神變真言經》不空羂索神變眞言經卷第十 《奮怒王品第十四》，高麗藏本
韩小荆《可洪音义研究：以文字为中心》p269
新集藏经音义随函录卷七 高丽藏
T1796_.39.0670b12
中华大藏经

According to the evidences provided by Ma Shijie, we can determine that U+2D223 𭈣 in the evidences is a variant of UK-30947, and the writer mistakenly wrote the ‘末’ as ‘未’. However, the annotations on the pronunciation of 𭈣 in the evidence clearly state that 𭈣 is UK-30947. Therefore, UK-30947 should unified with 𭈣 (U+2D223) and modify the glyph of U+2D223 𭈣.

As is agreed the right hand part of UK-30947 is clearly 沫 not 沬 so only if U+2D223 𭈣 is changed could they be unified, whilst SAT-04843 is the given source they may be other users of U+2D223 𭈣 dependent on the right hand part being 沬 and so changing would cause problems. A better alternative might be to encode UK-30947separately and then correct th SAT-04843 glyph and then move it as a second source to the new codepoint and the orphaned U+2D223 given UCS source or similar.

As for U+2D223 (SAT-04843), we find it semantically more appropriate to have 末 as the rightmost component rather than 未. At the same time, we confirm so far that glyphs with 未 and 末 have mixed appearance in older books, where correspond to the Taisho Tripitaka characters. Given the current state, we most probably keep the representative glyph and add an IVS that has the same structure with the UK glyph.