You are to assume that the raw voltages the load cell and strain gauge are spitting out are totally accurate and you trust the devices, so there is no source of accuracy error there. The strain gauge is actually custom made, so there are no specifications for that (that little square of wire loops that was part of the long column sticking down in the Instron machine).
In thinking about resolution, it actually has to do with what your analog-to-digital conversion hardware is capable of, which is the NI DAQ in this case. That device is 24 bits, so it's going to divide the range of output into 2^24 increments. However you also brought up the point about the digits that the LabVIEW interface actually outputs and saves to the text file. This resolution may be more coarse than what the ADC is actually capable of producing, so you should check just to be sure.
The statement of portability would be presumably based on the size of those devices and if they can fit in a field kit, as referenced in the scenario letter. Again, the strain gauge was the small square with the wire loops going through it, and the load cell is just one of those cylinders inside the Instron (you can see a picture of it on CTools).