Hi Fred.
I need to check my thoughts, but currently, they run along these lines.
As you know, the frequency of vibration ("pitch") of the note produced is a direct function of the weight of air being modulated. So, as with a trombone, you can lower the pitch as the tube becomes longer.
With a very large tube or with a very high pressure, the volume increases as the air gets compressed. So the frequency goes down.
The effect of steam is to add the weight of the water into the mass being modulated so the frequency gets lowered. But, the watter is a colloid (particles in suspension, not being dissolved, so somewhat independent). I'm not sure of the maths for that effect. Also, without knowing the air/water mix, it is difficult to estimate the difference in mass. But steam will weigh more than air at the same pressure and temperature, and thus have a lower frequency.
To further complicate it, the air is obviously heated by the steam and the hot gases occupy less space, gram for gram unless the compression is altered. So the frequency will increase when hot air air used.
I'd need a lot of facts to begin to work it all out. Best approach will be to make some "educated guesses" based on the recording when it was powered using compressed air. My initial thoughts were that a lot of these effects cancel out, so the recording we have may not be far away from reality.