I have found the problem at last!
So to recap:
Recap the Problem:
the whole website would lock up and become unresponsive on a frequent basis, messages to various devices that should have triggered didn't happen (like a Rule to turn a light on) ...
Discovery
I had set up a logic to detect whenever someone was at home:
I had LogicalRule with an OR condition for every PIR in the house, if triggered it set the Scene: "SC_SomeoneInHouse_On"
The Scene "SC_SomeoneInHouse_On" set a Switch Dummy Device "DD_SomeoneInHouse" to True
I had also created a DelayedScene "SC_SomeoneInHouse_Off", following "SC_SomeoneInHouse_On" that after 5 mins, reset the DD_SomeoneInHouse to False.
If any PIR triggered the Logical Rule before the 5 mins had expired, the Logical Rule triggered the SC_SomeoneInHouse_On again. So the DelayedScene restarted its timer.
The idea was that the DummyDevice DD_SomeoneInHouse remained True so long as there was movement in any room within 5 mins.
What I guess was happening was that as people moved about the house the SC_SomeoneInHouse kept triggering, setting up a load of background timers, even though on the DelayedScene I had "Restart timer on trigger instead of starting new timer" set to ticked.
Once I disabled the LogicalRule to stop the process everything settled down, the whole system returned to normal.
I am VERY so pleased about this because I have had this logical rule for a long time.
I look forward now to building on the network!