I am not sure wiring two of them and trying to subtract the 'Wh' readings is really going to give you what you want, and could actually be more confusing.
The issue with coulomb counting (ie integrating) power meters is that they integrate any small offsets or errors in the measurement, eventually leading to big errors and confusing data. The bi-directional meters have the advantage that at least their errors are the same in each direction - so if your shunt is a little off or the current meter is a little off it effects both the charge and discharge currents equally and the SOC measurement can still be fairly good. Using two separate meters, the errors are independent, so if the discharge current meter reads a little higher than the charge current meter, within a few days the difference will be telling you your battery is empty, even if it is full.
Secondly, even if they were dead-on-balls accurate, you would still end up pretty far off as they don't take into account charge efficiency (about 0.9), which means you need to put 1.1Ah back into your battery for every Ah you take out. Over the course of a couple of charge cycles this will make the math very tricky.
I don't want to be to much of a downer - the amazon or powerwerx meters could be very useful in understanding the power use by a specific appliance, such as your fridge, but don't expect them to be able to replace a dedicated battery power meter.