Backups are always "Full Zero level" but shouldn't be

Some context:

Avamar 19.1. Windows Server VMs (2008 R2, 2012 R2 and 2016).

Everything was running smoothly until we selected the option to "Exclude page file blocks when performing image backup on windows VM" and "Exclude deleted file blocks when performing image backup on windows VM". Windows Dataset configuration. We noticed that some backups where taking longer than usual and decided to revert to the old config. Therefore we deselected both "Exclude page file blocks when performing image backup on windows VM" and "Exclude deleted file blocks when performing image backup on windows VM" options.

After doing the above most, but not all, backups are taking a lot longer to complete. A lot means a lot. Several times more than usual. Looking at the logs we find the following which sheds some light on the matter:

2020-02-26 11:23:34 avvcbimage Info <11986>: Changed block tracking is engaged for this VM
2020-02-26 11:23:34 avvcbimage Info <11987>: Reference backup is #377 captured on 2020-02-26 10:50:11
2020-02-26 11:23:34 avvcbimage Info <14680>: A reference to a valid prior backup is available so this will be an incremental.

2020-02-26 11:24:55 avtar Info <8940>: Starting back up at 2020-02-26 11:24:55 WET as "root" on "avamarprx01.ptsuma.pt" (4 CPUs) [19.1.100-38]
2020-02-26 11:24:55 avtar Info <5731>: Not using include/exclude list

So far so good, however, a little down the log we find:

2020-02-26 11:24:54 avvcbimage Info <40659>: snapshot created:true NOMC:false ChangeBlTrackingAvail:true UsingChBl:true, ExitOK:true, cancelled:false, fatal: false
2020-02-26 11:24:54 avvcbimage Warning <43943>: Not an exclude backup but last backup has exclude list, do full backup directly to recover exclude contents.
2020-02-26 11:24:54 avvcbimage Warning <43947>: data exclude files changed,diskInfo setLevel0FullBackup(true)
2020-02-26 11:24:54 avvcbimage Info <43948>: A full level0 backup will be done when data exlcude file list changed.

And this happens again in all subsequent backups. So, it looks like Avamar, mistakenly, "thinks" that the previous backup contains an exclude list when it does not and keeps on performing full level zero backups.

I even deleted all backups of a certain VM, executed a new backup, which was, correctly, considered a Full Zero, as it, correctly, said it didn't find any previous bakcup. But the subsequent ones are always full Zero, throwing the logs that I quoted above.

This does appear to be a bug. Anyone has anything to say that might solve this, before I engage Avamar support?