I checked the system protection (shadow copy) option for the CloudDrive. It was already off. I am guessing that Windows by default sets this feature off for portable or virtual drives. I don't think this is the issue I am seeing because shadow copy would create the snapshot only at certain time intervals (or other somewhat predictable system event) . The excess activity I see occurs regularly during normal use of the CloudDrive.
I have done quite a bit of analysis and the problem seems to be related to the prefetch and pinning options. I can get the efficient CloudDrive behavior I expect by turning off prefetch and pinning of directories and metadata. There are other factors that seem to be involved here as well, including the drive sector size and storage chunk size. I will try to provide more feedback as my analysis continues.
The technical details window is not really that helpful. It shows the activity of commands and chunk ids, etc. but nothing that I can see equates this detail to the actual files or folder activity that caused the CloudDrive activity. I was hoping there was a way I could see such activity so I could determine what processes on my side are responsible for causing CloudDrive activity. A log that shows detail activity in a format readable by the user would be helpful.
Making the drive "read only" isn't a good option. I have already accomplished a similar scenario by changing the "Upload Threshold" for the drive to 10MB and the minutes to a high value (in my testing 600 minutes). This seems to keep the CloudDrive calmed down at least until the threshold has been reached. It would be helpful if the max bytes for the threshold could be set higher than 10MB. I will submit this with additional reasoning via the Stablebit contact system.