Whenever tacking a volume pot on the end of a circuit, or changing the value of the one already there, you need to examine the pot value in the context of all those other resistiors that might be in immediate direct contact with it.
The textbook case is the MXR Distortion+ and its cousins (DOD 250, etc). Some of these (certainly the one I borrowed in 1978), had a 10k output volume pot. The trouble is thet there is a 10k resistor in series with the input to that pot. True, there are diodes and a cap on the input lug of the volume pot as well, but we will confine ourselves to the resistor. From the pot's perspective, that resistor adds to the overall resistance of the pot. Essentially what you have is a 20k voltage divider (10k resistor + 10k pot) that is never allowed to go up past halfway. Even with the wiper at max volume, you'll have 10k on one side of the wiper, and 10k on the other, cutting the signal by half. If the pot is increased to 50k or even 100k, you end up with the equivalent of a 60k or 110k "pot" that is still never allowed to reach maximum, but comes closer to it from a proportional perspective. As a result of providing less attenuation, the output volume of those pedals where the pot value was changed is greater. Less attenuation means more signal level preserved/conserved.
So, changing volume pot value CAN, in some instances, change the potential volume/level of the pedal.
Is the Scrambler one of these? In a word, no. Let's take a look at the circuit and see why. http://www.generalguitargadgets.com/diagrams/scrambler_sc.gif
Imagine the 50k Blend control is set midway so that there is 25k of resistance on either side of the wiper. In tandem with the 4M7 resistor on the output, each half of the Blend pot serves as a sort of volume control for the signal. At the exact mid-point, there is a sort of voltage divider formed by R18 and each half of the Blend pot, with a total resistance value of 4.725k (4M7+25k). It is "turned up" pretty much most of the way since there is far more resistance in R18 than in what is on the other side of the R17 wiper.
So what would happen if you put a volume pot in there? Let's replace R18 with a 1M log pot (selected because it is a fairly common value). When the volume pot is set to maximum (wiper at the end where it is essentially in contact with the R17 wiper), you would have a voltage divider (which we'll call
R19) with 1M on one side of the wiper (i.e., the full resistance of R17), and somewhere between 0k and 50k (depending on Blend setting) on the other side of the wiper. If the Blend pot is set to the midpoint, you have 25k on one side of the wiper and 1M on the other; essentially a 1.025M pot turned down a bit but not that much. If the pot value was dropped to 100k, our R19 would behave like a 125k pot turned down about a 1/4 of the way, which would present a certain amount of volume loss.
So, in this illustration, opting for a higher value volume pot on the output provides for more signal conservation, as a result of its interaction with other adjacent resistors. Note that simply adding a 1M pot and leaving the 4M7 resistor in place will NOT improve matters.