Its hard to say who is right; Hugh or the alternate from the other vendor... One has the shunt on the positive side, the other on the negative.
I think I would start out with an experimental layout to figure it out for sure before I hooked it up to the intended loads.
Use a 9V battery to power the meter logic, and then use high value resistors (say 100k?) to probe the other connections, sourcing from the 9V. You'll find out which ones do what by how the meter reacts, without risking doing any damage.
I'd imagine you can tie the two negatives together (if they aren't already this way on the meter's board), and then probe the other two connections. One will make the volt meter respond, the other will go to the ammeter.
Do the probing without the shunt; you won't likely see any reaction from the ammeter if it's there since only about 90 uA will be flowing. You'll also be able to confirm polarity across the shunt this way, and determine which diagram is right.
One thing for sure, the original is dead off...
Steve