The specs on the CT-1 and its brethren are kinda rubbish, but then that's not unusual for a modern-day Behringer product (they bought Klark Teknik a good while ago) and not necessarily indicative of actual device performance. Still, I can't think of many more grateful mass production targets than this kind of mic booster. By the specs I'd guess it's a relatively simple circuit à la FEThead but presumably bipolar for cost reasons.
The sound coloration with the SM7B is likely down to a higher-than-usual 7 kOhm input impedance (which would be expected to result in slightly more highs and a more hefty bottom end). Should be less than with a FEThead though (33 kOhms or something if memory serves?). Not sure why the bottom end didn't seem to be affected much at all, the coupling caps couldn't be that small...?
The 25 dB maximum gain spec is a bit misleading - most people have mic inputs with 2-3 kOhms of input impedance, not 10k. I would expect closer to 13-16 dB of boost in practice. Still, that's quite enough to bring a mediocre input (say, -120 dBu) close to the physical limits.
As for those saying you need X amount of gain, well - that's neither completely right nor completely wrong. What you want is a given low EIN (like <-130 dBu shorted) and enough gain to safely bring this noise level at least 10, better 20 dB above that of the following recording chain.
Say you want to drive an ADC inside an average bus-powered USB interface, 3.3 Vpp (+3.5 dBu), dynamic range 104 dB. ADC noise is at -100.5 dBu, so you want a gain of about 50 dB.
Say you want to drive a mixer - dynamic range 110 dB, +22 dBu max. With noise at -88 dBu, you'll want over 60 dB instead.
EIN will generally degrade in variable-gain preamps when turned down substantially from max. There usually is little change in the top 10 dB, but if you intend to use a max 50-60 dB preamp to provide an extra 20 dB ahead of your regular mic pre, you may be in for a nasty surprise. In this case you would be better off using a 40 dB setting and attenuating the signal by 20 dB immediately afterwards, as silly as it sounds. Alternatively, use a purpose-built 20 dB preamp instead.
A preamp cascade is potentially useful if you're after minimum distortion. I have found typical compact mixer preamps to be happiest around the 25-30 dB mark. The lesser ones may have the 3rd harmonic come up to >0.1% or even close to 0.3% beyond the 50 dB gain mark (e.g. typical Behringer circuit), while at 20-30 dB they may hit around 0.003% even at highish levels.
The 0.02% THD at 0 dBu output (= -25 dBu input) spec of the CT-1 may not seem too impressive, but with a -59 dBV/Pa SM7B this level still amounts to about 126 dB SPL. Plenty of entry-level large diaphragm condensers make it to this kind of level at 0.5%.
