But about ten years ago I used L PADS they had a transformer inside
It was used to keep the load the same to the amp as you lowered the volume.
They fit in a reg light box where. Switch or outlet would go
I thought wirewound resistors had less inductance? All the non-inductive resistors used in my headphone interface were the non-inductive wirewound type. Although I do I know that air core inductors are simply coiled wires so it is something of a conundrum.
I would not be very concerned about inductance since speaker amps have to drive large inductive loads. The wire resistors inductance is a small faction of the inductance found in speaker crossovers. We do not need to fuss over this.
Is this the attenuation circuit you used? Was R2 8ohms and R3 2ohms?
Then you should have had about 14 dB of attenuation.... Wouldn't sound the same at all as connected directly across the amp output.
Now if you had the 8 ohm across the amp and the 2 ohm in series with the headphones, I could see the 8 ohm getting hot and the output to the headphone would only be 0.3 dB down with 50 ohm cans.
The 8 ohm is going to drop ~4X as much power as the 2 ohm no matter what you do with the headphones. Since we don't know if the volume setting was changed between the two configs, I'd guess the volume got turned down. A more careful experiment is required to clear up the confusion.