I did some measurements yesterday with and without a 15ohm resistor in parallel, and the only thing that seemed to change was the overall level .
The response looked very much the same, with a little flatter top end response when using the 15ohm parallel resistor.

I looked through your links to previous posts, and found the one where you suggest changing the l-pad and compensation circuit only. This is great news, and would make a cheaper and easier implementation of the drivers possible.
Also I will play around with Spice a do some simulations to confirm what I see and hear.

The Altec 902 driver sounds smoother and more natural than the PSD-2002, I will post measurement when I get them done.

Regards
Kim