This becomes tricky and a matter of taste more than anything else. For me, I like the "tip of the hat" to the old Fender thing where you have alnico PU's connected straight to a volume or balance control. In this case, the PU's load one another through the controls in a non-linear way that gives you that sort of "vintage"-ish vibe. It's a sort of crosstalk between the PU's that is very desirable to some people, myself included. If the bridge PU (for example) is full on and the neck is off...then you turn the neck volume up slowly, you're not just increasing the volume of the neck PU relative to the bridge, you're also changing the frequency characteristics of the bridge PU and, really, the whole circuit up to the preamp. If you could physically get the neck PU so it was not picking up the string and listen only to the bridge PU while you turned the neck volume, you'd hear some kind of change in the sound of the bridge PU.
The other BIG difference between vol/vol and blend/vol is...say the bridge is at 80% and you roll in the neck to 80%. The PU's are the same level as one another, but there's a bunch of resistive isolation between them which changes the way they interact. I can't say how, exactly...it's not something I've spent time investigating, but it's important to some people, and it's not possible to do in a bal/vol setup.
So in the Fodera preamp, there's no buffering before the balancer. If there were, I'd be using 25k or maybe even 10k pots to reduce noise since the buffer can drive that load easily. The PU's are passively connected to one another and then they hit the tone control...then the preamp input. The input impedance of the preamp is extremely high...over 1 Megohm...so it doesn't load any of the controls or the PU's. So when the bass is in active mode, the front end is still passive. The difference is that instead of having a wimpy passive output have to drive a cable to get to the preamp, it only has to drive an inch of PCB trace to get to the preamp input. Then the output of the preamp is put across the master volume. Then you need an output driver, or else you defeat part of the purpose of the preamp when you lower the volume slightly. Without an output driver, at half volume you'll have an output impedance of in the 100's of kilohms. Not good for noise etc. You could use a lower value (say 10k) vol pot which would at least get you closer, but then when you went passive you'd have major problems with that low value severely loading the balancer and tone control. If you buffer the PU's individually first, you have a much more linear situation. The outputs of the buffers SHOULD be linear and not affected by one another at one frequency more than another. From an engineering standpoint it's CLEANER to receive the PU's as differential signals and buffer them before blending. But my whole philosophy is to strike as nearly perfect a balance between "correct" process and real-world desireable result as is possible. It's not for everyone, but my opinions and practices come from actually doing the playing in the hot seat as much as from having artists tell me what they like. That's a big divide, let me tell you. No amount of endorsers' input makes up for knowing first hand what needs to be done. For me, it works the best, but I've also done it enough to be able get different results for different people who don't like what I like. In a Fodera bass, though, it's my (and Fodera's) vision of a bass preamp and we continue to refine and improve it.