`Submitter Names` lack spaces in `dataformat tsv virus-genome` compared to genbank file serialization

Describe the bug
The Submitter Names field has a not-very-robust serialization format of LAST NAME,FIRST NAME INITIALS,LAST NAME,FIRST NAME INITIALS... that does not separate individuals. Is this on purpose, if so why?

When I look up the original genbank file for a sequence, there is a space after the initials, before the next Last Name.

Compare output from

   datasets download virus genome taxon 186538  --no-progressbar  --filename results/ncbi_dataset.zip
 dataformat tsv virus-genome   --package results/ncbi_dataset.zip  --fields submitter-names

for e.g. OR084927 with what's shown for the corresponding .gb file.

CLI output: Kinganda-Lusamaki,E.,Whitmer,S.,Lokilo-Lofiko,E.,Amuri-Aziza,A.,Muyembe-Mawete,F.,Makangara-Cigolo,J.C.,...
Genbank file: Kinganda-Lusamaki,E., Whitmer,S., Lokilo-Lofiko,E., Amuri-Aziza,A., Muyembe-Mawete,F., Makangara-Cigolo,J.C.,

Note that the Genbank file separates names with a whitespace - which is prudent, as otherwise one needs to hope that the parity holds for long strings.

Reproduce the issue with the datasets download virus genome taxon 186538 and dataformat tsv virus-genome --fields submitter-names commands for OR084927, then compare the TSV value with its corresponding GenBank file. Trace the submitter-names serialization entry point and verify that names are separated as in the GenBank output, including a space after each initials sequence.