Here's my experience with the vmware 6 CBT bug so far:
We upgraded from ESX/vcenter5.5 to 6.0 in early October because my client got a new 6.0 upgrade license from Essentials Plus to Standard. Other than having to manually embed the HP b320i drivers into a custom ISO, it went well enough (drivers were not an issue on the 5.5 HP ISO).
It seemed like the backups were working OK after the upgrade. I would have noticed a 7 hour incremental vs a 15 minute incremental. In early November we set up a cluster and HA in vcenter (it was a simple datacenter setup before). After the cluster setup, I started getting CBT data is invalid on 4 VMs every time the backup ran. I tried running fulls and manually clearing out the CBT data. Sometimes it would fix the issue for a couple days on those VMs, and sometimes not. The issue would keep happening on those 4 VMs (out of 13 VMs).
Last night I downgraded the 3 hosts to 5.5, and left vcenter at 6.0. Reconfigured Veeam and the backup jobs. The new active full just finished a little while ago and I ran an incremental. There was 1 VM that had CBT data invalid, and it was the Veeam VM that I converted P2V at the same time we set up the cluster, so who knows what the root cause is exactly. I am going to remove the Veeam VM from the backup set and run a few more incrementals to see what happens. I'm pretty cranky about it because it's totally screwing up the backup window and the offsite copy job.