Doesn't that still cause you trouble, with the guitar's internal witing? Tone pots, pickups, etc?
Less so if the load isn't reactive. Let's take a pickup, again say 10K output impedance. If we use our 1:10 rule for source:load impedance, we can get away with a 100K load without affecting tone. What? All guitarists immediately object! But it's true!
You see, the pickup in a guitar is subject to several loads: volume pot, tone pot/cap, cable, amp input impedance. Let's take the tone pot. Say we have a 22nF cap. That has a reactance of about 70K at 100Hz, 7K at 1kHz. So practically speaking, our pickup has three loads at high frequencies: 250K volume (or 500K maybe), 250K tone, and 1M amp (some amps are less) PLUS the cable capacitance. Let's say that is 1nF. That's really the biggest problem; reactance of the cable at 10kHz is 16K; otherwise we are still pretty close to 100K net parallel load.
The next problem is what happens when we use the volume pot; I discussed that already but basically it increases output impedance. That makes the cable capacitance problem much worse. So eliminating that load from the guitar must be the first and most important goal.
Note that some guitars have their volume knobs wired so as to load down the pickup (in order to have the volume controls for each pickup independent of the other). The pickup itself has a complex impedance, so that will change tone for a different reason, since the voltage divider formed by the pickup and volume control will become increasingly variable by frequency as the volume control's load gets closer to the pickup's nominal output impedance.
The DI can't solve that problem, so . . . if you want to preserve tone, you don't do that. But how to solve the pickup/volume control problem? Well, you can add some resistors to buffer the volume controls from each other, somewhat anyway. That does increase output impedance, but since we've eliminated cable capacitance that is not too big of a problem.
It's tough if we want to stay passive. I came to the conclusion that the only way to solve (most of) the problems of guitar wiring in a passive circuit was to use a step-down transformer (hence, the DI right at the guitar's output). That means we need an amp with sufficient gain, basically a mic preamp. That really isn't too hard. It would be nice to have a built-in transformer that is specifically designed for the task at hand though.
The other way is to go active, and buffer the pickups and each control from each other. So guitars could run off battery, or switch to phantom power. That works perfectly, but requires power.
