I haven’t done this, so I guess I’d be interested to hear how it goes. I don’t think it’s a bad idea to just try it in batches of 50 and see what happens. The worst case scenario (in my mind) is that the queue fails and there are WordPress records that get deleted without the plugin deleting their corresponding Salesforce records. If it does it would be good to see if you can tell why it fails and if it’s indicative of a problem.

I don’t think a temp SF ID would affect the batch. What the plugin should do is batch up all of the relevant requests and run them in succession, but it doesn’t stop running if one of the operations fails. It would just create a log entry.

One thing you could do, though, is check the plugin’s Mapping Errors tab (/wp-admin/options-general.php?page=object-sync-salesforce-admin&tab=mapping_errors) and see if there are records there. If there are, you could delete those mapping object rows and this would reduce the number of unnecessary bad sync records the plugin tries to process.

In any case, I think you can do whatever you think is best and see how it works. I think if it were me, I would export a list of the WordPress ids before deleting them. Then you’d have that list, if you ended up having to delete them from the object map table, and correspondingly you could get the Salesforce IDs from that table before deleting them to delete them from Salesforce. I think the plugin might be fine for doing all of this without issue, but it’s not really built for that purpose so I would want a backup.