You Reply: Beginning to get somewhere with this finally! I've only managed a couple of tests, but the current draw in deep-sleep seems to depend the input voltage..
Using a 3.3v supply I get a deep sleep in the region of 35uA (both using vcc and 3v3 pins as power in)
But with an input voltage of 3.0v I get something in the region of 700+ uA
It seems strange to draw more with just a 10% drop in supply, but in a couple of tests its been repeatable.
I was originally testing with a moderately discharged LifePo4 battery, with a full charge battery the voltage is ~3.3v and the deep sleep current is low. I'll now have to use it a bit until it drops to its stable voltage which is closer to 3.2v and see what happens.
Will keep you posted