Limit concentrations of client requests to spread out server load

[wizkid057]

Server side load gets momentarily concentrated unnecessarily for several rounds of template updates after a network block change.

The network block change load is expected (currently... future enhancements may make this a non-issue). However, with many users on the default work update interval there continue to be load spikes for several rounds of work updates concentrated around the default work update value.

These concentrations continue until they eventually diverge naturally, or until there's a new network block.

Probably something as simple as a random -5,000 to +5,000 millisecond delay delta on the first work update after settling in on a new network block full work would be sufficient, provided the random value is executed with sufficient temporal resolution.

To be clear, this doesn't adversely affect anything currently, but an enhancement to help with future server side capacity scaling is certainly best long term.