Personally I’d really like to see a high throughput processing function.
Currently, when I prepare a draft I always put temporary citations in the text in the format of [XXXXX], where the string/number is specific to the reference (e.g. PMID). In the last step, I convert these temp ref to endnote library manually by searching these numbers and insert one by one, which really is a whole day of work that should be automated.
Have you tried (does EN20 allow) you to change the field for the EN temporary code from record number to accession code or label field.
You can also change the temp citation delimiters from curly to square brackets there if you prefer those (these options are under “temporary citation” in preferences).
You could copy the PMID number for your library to label field (Endnote tools, change/move/copy fields in endnote) regularly or make sure your connection saves the PMID to the label field for new records.
Then turn off the preference under formatting to “Omit Author and/or Year from formatted citation if removed from temporary citation” and endnote should populate the citations correctly.