Colour films involve tightly controlled processing, therefore there is only a really limited number of actual variations on processing, which explain the successful results coming out of shooting at box speed.
On the other hand, with B&W you can get usable images through so many different means, that it is why we must understand that ISO ratings are meaningful and accurate only in an ISO situation.
We all know that using a given developer allows us to shoot at EI #1 and using another developer requires us to shoot at EI #2. There is no trickery here. If you cook a leg of lamb at 225F you can do so in seven hours, but if you cook it at 350F you should be OK at 2h max.
The standard for ISO film speed involves a particular developer, processing sequence, and measurements. The point is not to have something that is perfectly representative of the average consumer experience, but to have something that is controllable in production circumstances, batch-to-batch consistency among other things.
If you want to verify the true ISO speed of a film, get a copy of the ISO standard, and do the testing accordingly. Otherwise you're just reflecting the bias of your experimental circumstances.
Snapshot, none of the developers you are using are part of the ISO standard. Why would you expect your results to match precisely this standard?