Appears that the Sonoff POW R1 has issues in measuring power and that you will see spikes in the reported MQTT messages. For anyone requiring Power measurements that may use HS events to trigger on those values, get the R2 unit. The measurements are offloaded to a seperate internal device and reported to the ESP8226 chip rather than it calculating the values. With the recent codebase changes going from 5.x to 6.x tasmota, the power readings are erratic as more CPU time is being used to service "other code". I raised an issue and ultimately Theo advised to go R2.