Sadly they only ever implemented data deduplication across a single machine's backups. They did promise the feature of account wide data deduplication, but as far as I'm aware they never delivered on it.

So our inactive old backups really are taking up extra space even if they largely duplicate other machines.

In my case it is actually CrashPlan's fault that one of my two inactive backups exists. Their mechanism for migrating a backup from one machine to another used to be well known to not work well. So when I replaced my home server hardware a couple of years ago I didn't use the backup adoption feature and just started a new backup.

I moved data across from the old server to the new one, reorganising/tidying it as I went. I kept the old backup around just in case I had missed anything in the process. As yet I haven't spotted anything I missed, but the whole point of that backup is, as a, backup...

That blog post I linked to suggest that if you contact CrashPlan support and explain that 5 days isn't enough warning that they'll probably give you more time.

In my case I just cheated by using a couple of VMs to refresh those "inactive" backups. I'll delete them at some point, but I'm certainly not going to rush around dealing them them now just because CrashPlan changed their terms, didn't tell* me about it, left it for 3 years and then gave me 5 days to sort out the issue !

* maybe they did somehow indicated they'd updated their terms to the detriment of the users, but I haven't found any user who knew about it yet