The useful info in that is to determine the amount of energy produced by the lamp in different spectrums. The 575W produces nearly the same visible luminous output as the 750W lamp. Knowing that to produce the same light that filaments must be hotter to produce a higher
efficacy we also know the color spectrum will
shift to a higher temperature.
With that insight we can also postulate that the spectrum of the 575W is shifted towards the UV end (more UV production) and the 750W is more towards the IR red (more IR production.)
Now since the 575W is producing more of the input energy in visible light and less in other ranges both IR and UV compared to the 750W lamp we have to determine wether the visible spectrum (which color filters affect the most) degrades color filters or if non visible spectrums do.
Having a higher
efficacy the 575W lamp produces much less non visible light than a 750W lamp of the same luminous output therefore the 750W lamp will produce UV and substantial amounts of IR energy where as the 575W does much the same but lesser amounts and shifted towards the UV energy levels.
My general assumption is that due to the fact that 100% of the energy produced by the lamp ultimately is converted to heat (light and heat) that the more input
power you have the faster you will degrade your color filters. Under that assumption the 575W is the way to go unless you have a very specific reason to use extra long life lamps. Gobos and other light absorbing accessories will also benefit from lower thermal absorption.
Another advantage to 575W lamps is the lower
power requirements; that equates to lower
instrument temperatures, longer life in all accessories in the
instrument, lower heat into your
venue, and the ability to use more instruments on a single
dimmer for washes and other
stage filling lighting uses.
The downside is lamps will burn out more often. Also check the
color temperature difference in the lamps, I expect they are a few hundreds degrees different minimum.
Truth be told all this typing didn't scientifically prove that one lamp was superior in extending
gel life.