Thank you for writing to us. To your question, here is the explanations.
as per the design of UDP, In custom recovery point retention, there should be at least one to retain, due to the fact that UDP enables you to run custom/manual backups, and incase of any custom backups that you run, it needs such a configuration to keep that backup recovery point. Since it differentiates the recovery points in terms of Daily, weekly, monthly and custom, There should at least be one in there, so when you run a manual backup, it can retain it in that one place.
To get rid of that one old custom RP, you can run another custom incremental to take its place, but if that custom RP is a full backup and is the base line to the incremental backup chain(all the other backups), it always will have that data in datastore as every other sequential incremental are referring to it.