i am not sure what do you mean by bit error probability. But to measure energy usage you can use an resistor and osciloscope connected in serial connection. We did this for other WSN nodes, you can find more datails in our paper "Realistic Model of Radio Communication in Wireless Sensor Networks".
Digital system delivers pulsed power and normal power meter measures average power. Use peak power meter.
Set the maximum bit rate as an input data stream and measure power at antenna port. Convert peak power into average power on a time scale by ratio of pulse ON to pulse OFF time in a second. Measure DC bias voltage and current and calculate DC power taken by module. The difference between DC power intake average and output power at antenna port is the power consumption of the modile
nRF series is developed by Nordic semiconductor. Usually I found their data sheets quite descriptive of energy consumption in different modes. Power consumption depends on application specific profiles too. While oscilloscope can give you rough estimates, precise characterization will require specialized equipment.
Depends on the type of node. If you have an external device like Agilent 66321D DC source. |Then you can measure the current very accurately. And derive Power and Energy using this value. (maybe Oscilloscope or multimeter also but they will be less accurate than Agilent) . Else you may try to sample the ADC. However ADC method will not be as accurate as the others
Depending on the system you are using, you may discover it is possible to query these information, especially if you have an application that allows you to remotely monitored your nodes in the network. This may prove to be efficient since it would be done remotely. I think such is possible with the WaspMote (see http://www.libelium.com/products/waspmote/ just to have an idea).