Help to interpret Accelerometer data...

Is it just custom fitField with corresponding data displayed in Monkeyraph

Could someone please clarify how to get these visualization graphs from the first post of this thread? Is it just custom fitField with corresponding data displayed in Monkeyraph

I am looking at using accelerometer data in a weight training app and I thought I would re-open this discussion as it seems to be the repository of most of the useful information about the accelerometers.  Due to sparseness of the formal documentation on the accelerometers I am making the following assumptions:

1. the accelerometerData is raw data from the in-built accelerometers sampled at the rate specified in the Sensor.registerSensorDataListener method.  This means that the data first needs to be put through a low-pass filter designed around the sample rate. Usually this would be a FIR filter which would introduce a phase delay, which can be removed by using a function similar to matlab's filtfilt() function.
2. the devices do not have any way to measure power – they would need to have some sort of pressure pad and you would then have to use the device to apply a force to something in order to be able to do that. Note that this is quite different from the Garmin Vector 3 bike pedals which I imagine do have pressure pads or equivalent to measure the force applied to the pedals/crank.
3. According to my understanding of dynamics, power is the time rate of doing work, or F . v (the dot product between the Force doing the work and velocity both of which are vectors). It is measured in watts with dimensions of [Mass]*[Length]^2/[Time]^3 and cannot be the same as acceleration which has dimensions of [Length]/[Time]^2. If the power values are indeed the magnitude of the accelerometer vector as mentioned below, then this is a major misnomer. 
4. As mentioned, accelerometers also measure the acceleration of gravity and for the data to be useful this would need to be removed based on the orientation of the device within the gravitational field. I assume that this is not done in the calculation of the 'power' value as the orientation of the device in the gravitational field would not be known. As we don't have a gyroscope or equivalent stable orientation reference it would require a sophisticated calibration step when the device was stationary and by monitoring changes in orientation post calibration.

any update on those tickets? can you link the original ones?

An old post, but still relevant.  I've just noticed that the accelerometer axes on the Garmin Edge Explore 2 are not like other devices.  

+X axis extends from 9 o'clock on the watch face, -X from 3 o'clock
+Y axis extends from 6 o'clock on the watch face, -Y from 12 o'clock
+Z extends from the face, -Z from the back.

Essentially the Edge Explore 2 is (-x, -y, z) compared to other models.  Does anyone know of any other devices that are like this?