DOes the colour temperature change with mercury lamps as the pressure increases? and/or the efficiency?

If we assume for now none of the glass and eventually air absorbs any radiation, the efficiency (radiated vs input power) of radiating power decreases with increased pressure.
But when speaking about an efficacy (lumens vs input power), this tends to increase with higher pressures.
The reason is, the primary mercury emission lines, are qround the 200nm, but these get selfabsorbed by the higher mercury pressure around. With low pressure the efficiency uses to be above 60%, that is why fluorescents became so efficient (even after phosphor conversion into visible, the efficacy reaches 100lm/W for a high quality light, what means an efficiency in the 30% range)
But the 200's nm is vacuum UV, not visible. For visible output you need an already excited atom to get excited even further, only then the radiation may move to the visible. And for that you need high percentage of the atoms to be excited inthe arc stream, so high arc loading.
And that loading increases with the mercury pressure. There the natural selfabsorbtion of the primary lines converts that energy into heat necessary to mwintwin the arc, so more energy is "freed" for the visible light generation.
So the higher pressure means higher arc loading, that means more visible light radiation from the higher excitation states (so longer wavelengths, so more visible an more of the longer wavelengths, so better color) and because of the higher temperature, the Doppler effect from the faster shaking (= higher temlerature) atoms on the radiated light causes the spectral lines to widen (pressure broadening), so improving the color as well.

If we assume for now none of the glass and eventually air absorbs any radiation, the efficiency (radiated vs input power) of radiating power decreases with increased pressure.
But when speaking about an efficacy (lumens vs input power), this tends to increase with higher pressures.
The reason is, the primary mercury emission lines, are qround the 200nm, but these get selfabsorbed by the higher mercury pressure around. With low pressure the efficiency uses to be above 60%, that is why fluorescents became so efficient (even after phosphor conversion into visible, the efficacy reaches 100lm/W for a high quality light, what means an efficiency in the 30% range)
But the 200's nm is vacuum UV, not visible. For visible output you need an already excited atom to get excited even further, only then the radiation may move to the visible. And for that you need high percentage of the atoms to be excited inthe arc stream, so high arc loading.
And that loading increases with the mercury pressure. There the natural selfabsorbtion of the primary lines converts that energy into heat necessary to mwintwin the arc, so more energy is "freed" for the visible light generation.
So the higher pressure means higher arc loading, that means more visible light radiation from the higher excitation states (so longer wavelengths, so more visible an more of the longer wavelengths, so better color) and because of the higher temperature, the Doppler effect from the faster shaking (= higher temlerature) atoms on the radiated light causes the spectral lines to widen (pressure broadening), so improving the color as well.

if you can just increase pressure to improve both efficiency and light quality why have they not done so to compete with other lighting sources? was it impractical for lighting manufacturers or dangerious or both?

In one of my older projection handbooks they show a British 35MM cinema Gamont-Kalee projector using a water cooled HP mercury lamp in the lamphouse.Fast forward to today-Sony was using HP mercury lamps in their digital 4 K HD cinema projectors.Sony just got out of the cinema digital projection-so theaters are slowly replacing their Sony projectors with Christie,Barco, or NEC.These use xenon lamps.

if you can just increase pressure to improve both efficiency and light quality why have they not done so to compete with other lighting sources? was it impractical for lighting manufacturers or dangerious or both?

If you boost the pressure/arc load, the efficacy saturates. The reason is, the broadening starts to extend into invisible parts of the spectrum, so the energy becomes lost again. So the maximum attainable efficacy is about 60..70lm/W (for really huge power and UHP lamps). The light quality rises to nearly 100 already with the UHP projector bulbs. But these are designed not that much for the overall efficacy, but for the point brightness, required for good system efficacy of the overall projector system.
The problem with huge power levels needed to reach the efficacy are the losses from the support materials absorbing the light (the arc becomes very small, what means in fact hidden within the electrode assembly).
Plus the high pressures and loads mean so high thermal load, the known materials can not stand it for too long, so the lifetime is rather limited (few 100's hours for a typical projector bulb).

Yep digital projection uses UHP lamps in their conventional projectors. They will use 1-4 lamps in a projector depending on the brightness needed. They are rated for 1500 hours but I usually replace them around 750 as they start getting red heavy as they age and start to throw off white balance. They are also known to fail in quite a spectacular manner at times. I have cleaned glass out of projectors on many occasions. In rare cases they have caused damage as well that required units to be repaired. I've had to clean up xenons that have popped in some barcos also.

If you boost the pressure/arc load, the efficacy saturates. The reason is, the broadening starts to extend into invisible parts of the spectrum, so the energy becomes lost again. So the maximum attainable efficacy is about 60..70lm/W (for really huge power and UHP lamps). The light quality rises to nearly 100 already with the UHP projector bulbs.

The broadened mercury lines emitted by UHP lamps are very prominent in the emission spectrum while the red output that arises from Hg-Hg molecules and electron-ion collisions is still too weak. As a result the light color quality is not even remotely close to 100 Ra8. The measured light color properties of a 150 W UHP burner (200 bar mercury pressure) are: 8622 K, 73 Ra8. However, the spectrum is changed by the dichroic mirror used in the complete reflector lamps, resulting in lower CCT of ~7600 K and CRI of ~60 Ra8 (these figures may vary with the lamp model though).