Your 'estimate' is almost certainly wrong and is based on your conclusion then worked backwards to 'prove' your LSD attempted point.
The error margin is ±20% at most, the estimate is definitely not an order of magnitude off. There is no working backwards or forward, either the number is right or it is not.
The calculation is based on the resistance per square (the seller lists the PCB copper weight as 1oz./sqft). Based on the shape of the copper (counting pixels), I've calculated the resistances between the 4 junctions and from that, I've calculated the portion of the voltage that the sense terminals (right side) see when current is injected into the force terminals (left side), resulting in the resistance observed by the meter.
You can start a new thread if you feel that this question needs a deeper dive (or if you feel like your hunch about my calculations is almost certainly right), I'll happily contribute. I'm done derailing this particular thread for now.
Last time there was a THREAD on this I remember it was in the 10's of nV with this sort of layout well below the LSD on a 6 1/2 digit meter.
The thread you linked talks about that number in regard to a bent solid copper wire short. That wire is also shaped in a way that current between the force terminals doesn't flow by the sense terminals, such as in the product that you've found.