One of the LL1674 seems to be off-spec, the secondary static resistance should be about 605 ohm and 3 ot the transformers do indeed measure 600-608 ohm. The fourth one measures 560 and 600 ohm for the two seconday resistances, this would lead to an off-balance surely?

Gain matching and CMR would not be so good.
Difference in copper resistance is due to difference in wire gauge. It is possible that we used wire from different runs or from different vendors. However, the number of turns should be correct. (I say ”should be” as shit do happens, even at Lundahls. But if the number of turns was not correct, the difference in copper resistance should have been around 100 ohms for the LL1674 (representing two full layers).)

Which problem will arise? The difference in DC resistance of the transformer should be compared to the impedance of the load. If the load impedance is 20k + 20k, (reasonable or even low load for the LL1674), the 40 ohms difference of the transformer is in the magnitude of 1/1000 of the load impedance. In my opinion, this is ignorable. Besides, is the core degaussed after measuring DC resistance with a DC current? This is probably a bigger problem.