Charge them up fully then take the measurement. Ir is supposed to be taken on a full cell. It goes up as the cell is depleted. And 3.4 volts the cell is almost empty and has a button top added which will make it go up some. All my button tops have higher ir then there unprotected siblings. They maybe fake but taking the measurement at full will tell a lot more. If it’s still that high at full I’d start a dispute but it should go down
Despite being an industry standard, to me AC internal resistance testing can be used to track cell health but not much else. Also, rail contact resistances matter (a lot). I've being tinkering with my Lii-500 and yes, rail resistance can vary a lot between insertions. However, since charger measures IR on startup it is easy to do a few cell re-seatings until we identify the lowest values (which I'd consider and average instead of averaging all obtained figures) which show up when rail contact resistances are at minimums. This also matters for cell binning since a high contact resistance on a given rail put that rail's cell at a disadvantage versus its neighbours: the higher the rail resistance the less energy/charge will go into the cell (because of the cut-off current value) and the higher will be the actual cell voltage cut-off because of a larger delta between rail and cell voltage (dV = Icharge × Rrail). 
Cheers 