If the normal system is saying there is an error, you likely want to stop this attempt to free space. Something is wrong with the system or the filesystem so you want to figure that out if your backups are at risk.
In the past with HFS only systems, the bypass command allowed someone that was sure about which file or folder needed removing to surgically perform the same “delete all versions” of a file that the gui allowed.
Running that on one file (rather than a folder) might be useful to see if there are detailed errors that your system didn’t print when you tried this delete.
If you need any files that were backed up, it’s safer to get a new drive to start a new backup. Then go over the old drive and decide how much of it you need. Just a few files could be hand copied back to the main system, backed up and then deleted. Simpler is to just label the old drive “backups from 2015 to 2022” and archive it on a shelf. Hopefully it will serve you until you are sure you don’t need any data and then you securely erase and put it back in service or recycle it.
Some unsolicited advice, The benefit of backups is you don’t have to know what was lost to get it back later. I wouldn’t risk that if I could safely spend $100 on a new 4 TB drive (for instance). It’s hard to guess your setup though - we don’t know the value of your data, time and budgets for getting more storage or spending time learning why the delete function failed…
bypass
command. Also this be sure to read the comments. I don’t use bypass very often as it’s very dangerous if you make a syntax mistake. I usually just bench that backup drive when it’s full and start a new one with the old drive excluded