Then it must be a clocks issue - please, install the NetTime and sync your clocks.
Regarding low usage. As you already knows - the network is used by real people, not machines, thus no “normal” or “abnormal” usage - any customers’ behavior is normal, we can’t use the past usage to predict a future usage.
Back to possible technical problems.
Since you uses a VPN, it’s expected to have greater latency than a direct connection, so your node could have more cancelation of uploads than without a VPN. You can use a success rate scripts to check this:
The second - since your VPN provider have a limited number of public IPs, there is a high probability that someone else running their storagenode in the same subnet /24 of public IPs, all such nodes are treated as a one node for uploads (i.e. only one node is selected from the /24 subnet of public IPs for each piece of the segment of the file), thus all such nodes shares the traffic, i.e. two nodes would have only half of the traffic each, four only quarter and so on.