Logically, switching between router and modem mode won't make any difference other than the (perhaps temporary) benefit of a restart and re-registration of the hub, since if there are any latency problems on the VM network these will be between the cable modem to CMTS. I'd restate my view that on the figures originally posted by @umar328 there's too many T3 timeouts across all upstream channels, which should not be there, based on the implied run-time from the pre-RS levels on the downstream.
We don't know if there is a VM latency problem yet, so something else for you (umar328) to try is to setup a Broadband Quality Monitor. That will show if anything is going on with your VM connection. Post a LINK to a LIVE, SHARED graph here and we'll see what's happening, don't post screenshots or a link to your logged in view of Thinkbroadband. Usually needs to run for 24 hours before we can draw reasonable conclusions, but the live graph will continuously update so you can post the link immediately.
Umar328 - sorry if this all sounds a bit techno-babble, bear with us, setup the BQM, usually these sorts of problems can be resolved. The other (main) possibility is that there is a latency problem due to any work VPN routing, but that would not normally affect Zoom calls if that's the "public" service.
