dddd

Sorry that I didn't bring this up earlier, but I've now thought about why heating a stem works so well at loosening it's bond inside of the steerer tube, even as the aluminum part expands inside of the steel tube which expands less than the aluminum.

What actually happens is that the aluminum part lengthens in response to the heat, and because of the difference in length-wise expansion, the stem breaks itself free of much of the length of the corrosion bond, with perhaps only a finite length along the bond's length remaining stationary and thus still bonded.
I think that this explains why my Centurion Pro-Tour's stem broke free with so much less twisting torque after I had heated it but while it was still hot and expanded inside of the fork's steerer tube.

With your stem quill now cut off (to what length I don't know), is there still enough length protruding with which to apply heat and twisting force?
Perhaps two or three smoking-hot heating cycles (including dry ice after each heating) might allow a limited amount of torque to still break it free.

If it's near-flush with the end of the steerer, then I would first drill it to as large of a diameter as possible.
A 7/8" drill measures 22.225, which is likely no bigger than the steerer's ID. And the very outermost edge of the tip of the bit could be smoothed to prevent any possibility of the bit digging into the inside of the steerer. I don't really know what progression of drill sizes would be best to use, but the bits themselves will be quite expensive if you don't already have them.