Fix Monitoring2Monitoring blob name generation, handle size mismatch when checking if blobs are synchronized

[joelverhagen]

This PR fix two bugs I've found in dev based on SDK move (1 and 2, below), and one pre-existing bug found due to blob corruption in DEV (3, below)

Summary of changes:

1. Scrub blob URL in logging, so we don't log SAS (query string)
2. Handle mismatch of scheme between base URL (http) and blob item from list blobs (https)
   • This difference was leading to a substring being off by 1 ("http" is shorter than "https")
   • It would be a much larger change to make it so we never use "http" in memory, this behavior pre-existed in the SDK move (example)
3. When blob sizes are different, do not consider them synchronized.
   • In DEV, there were some old blobs that had matching SHA-512 metadata but they were actually different. This was causing catalog2dnx to no-op. This can be easily avoided by also checking the file size. Of course this doesn't fix all problems if the SHA-512 is different, but it's an easy, safe improvement.