The 6HS5 and similar tubes are rather unusual. They have very high gain (comparable to a sweep tube, which should become no surprise), yet very low perveance: about 100mA plate current at Vg = 0V, Vp = 1000V! I broke one open long ago (these pictures are actually dated 10/2003, just seven months after my website opened in fact; it is now 2008!) for inspection. These are the results:
You can see the top mica is even punched for a screen grid, so this seems to be no more than a stripped-down sweep tube. The cathode size, grid dimensions, beam-forming grid and plate are identical to compactron sweeps of the same period. The high plate resistance (about 10 times that of a triode-wired sweep) is due to the low electric field reaching the grid, which is due to the wide distance between the plate and grid, as compared to the much smaller distance between grid and screen in a beam pentode.
Possible uses? Who knows: zero-bias class B RF final; electrostatic loudspeaker driver; high voltage pulse regulator (wait, that was its original use!); 6.3V, 1.8A resistor; etc.