DCIR changes with state of charge yes, it is significantly higher when SOC is low, slightly higher when SOC is high ( See the last graph )
It’s also higher when cold, lower when hot.
In his review HKJ found that the DCIR function is reasonably consistant but that it significantly underestimates it.
For reference Docware measured the DCIR of several cells according to IEC61960 standard:
But even if the S4 underestimates it you can still use it to track ageing.