Feature/improved_genomeID_resolve

[liannette]

This pull request updates the genome ID resolving process in PODP mode.

It builds upon pull request #312 and introduces the following improvements:

• If no valid RefSeq assembly is available when resolving the genome ID, a GenBank assembly accession is attempted instead. This allows antiSMASH jobs to run for genomes that have a GenBank accession but no RefSeq accession.
• When a genome record has a RefSeq ID, it is verified via NCBI to check its validity and determine if a more up-to-date version exists.
• The genome status JSON is updated after each attempt to retrieve BGC data—whether successful or unsuccessful—instead of only after processing all genomes. This helps preserve progress in case the process is interrupted midway.