Personally, I think it's too soon to blame possible custom hardware inventory. Because it looks like, at least for this specific machine and this specific log, it can't even do heartbeat ddr:
"<![LOG[Inventory: Opening store for action {00000000-0000-0000-0000-000000000003} ...]LOG]!><time="16:27:04.146-300" date="09-28-2020" component="InventoryAgent" context="" type="1" thread="1380" file="datastore.cpp:176">
<![LOG[CInvState::VerifyInventoryVersionNumber: Mismatch found for '{00000000-0000-0000-0000-000000000003}': 2.0 vs. 0.0]LOG]!><time="16:27:04.224-300" date="09-28-2020" component="InventoryAgent" context="" type="1" thread="1380" file="agentstate.cpp:1499">
<![LOG[PutInstance failed on 1 attempt. HResult: 80041013]LOG]!><time="16:27:04.260-300" date="09-28-2020" component="InventoryAgent" context="" type="2" thread="1380" file="wmiutils.cpp:322">"
I suggest first confirming how many devices have stale heartbeat, hinv, or sinv. I forget when it was added to the console; hopefully it's there in v 1810. In your console, go to "monitoring", then "Client Status", then "Client Health Dashboard". one of the charts is for "Scenario Health"; which lists how many are successfully doing multiple client actions, like policy requests, heartbeat, hardware inventory, and software inventory.