feat: update_manager - use a mutex

[JoeSchubert]

currently due to the async nature of these calls, the update_manager is not creating a once/second que as expected. Therfore, we should acquire a mutex to lock the process in order to enforce a bottleneck here

[JoeSchubert]

This should go into beta first for more people to test. In my testing, it works and limits the API to queries to about 1/second as intended. However, I would prefer for a bit of a wider test before simply rolling it out.