You'll find a good description (though slightly out of date with the present market) on how it all works here:
http://bicycles.stackexchange.com/questions/5612/how-do-bicycle-power-meters-work
What is being called (none) linear is really about the impact of force on the pedal, crank, BB and how it deforms (twisting/bending etc) these components, so to get the full power output you need to measure all the vectors the force acts through all the way through the pedal cycle, clever alignment of strain gauges etc help ( because of the way they work,
https://en.wikipedia.org/wiki/Strain_gauge ) but it's perhaps not economically or technologically viable to cover the object in enough orientated gauges to capture all the vectors of forces in action all the time, so a compromise has to be made with some clever maths to fill in potential gaps.
The key thing as pointed out by them that has 'em is consistency, regular interpretation / tracking of data trends and testing to establish a baseline (FTP / power at various durations).
There's plenty of guides out there on how to use power meters effectively either for improvement or pacing etc
Battery life with Vector 2 has the same issues as Stages in that it is very effected by temperature, in the winter with temp <7 degreesC I was lucky to get a 200km ride out of a set of batteries sometimes. Presently I am getting over 1000km on a set of batteries. I also suffer from occasional inconsistency in balance between left and right measurements - the root cause of which is often obscured as anyone who has any Garmin product will know, we are all their beta testers! LOL