Thanks for that link, Mike. You always come though!
Resistors and diodes, huh?
Not exactly a DI, and notice (according to the author of this article) that the instructions say to plug it into the high impedance input of your interface:
"...but in the studio you'd plug it directly into the high-impedance instrument (not line) input of an audio interface." So if you don't have a high impedance input on your interface, you'll probably want a conventional DI.
There's also this:
"...which is designed to emulate the impedance of a dual-input amp (probably a Fender Twin)."
From a Fender Twin owner's manual:
Input 1, both channels: 1M ohms.
Input 2, both channels: 136k ohms.
And this:
"Measuring the input impedance of the ZBox while it's disconnected from the audio interface shows the high-impedance input to be in the order of 300kΩ and the low-impedance one about half that." He should have measured it connected. This will have some impact on the overall input impedance but without a schematic it's impossible to know how much. As for the Fender emulation, it looks like the low impedance input is close to Fender's Input 2, but the high impedance input seriously missed the mark.
"This is achieved using a network of resistors, but the design also includes a chain of diodes that goes some way towards simulating the way in which a valve-preamp stage reacts to signal level. The diodes are arranged so that they can't cause clipping at normal pickup output levels, so their effect is mainly on the input impedance, which will reduce as the input signal level increases." I'm guessing a tube guitar amp's input impedance shouldn't significantly change with input level if the circuit is designed well. I'm thinking the diodes are for effect and not meant as an emulation. (speculation on my part

)
"There are no capacitors, as the guitar cable provides the necessary capacitance, but for use with a radio transmitter (where you don't have a long cable) a switchable cable-emulation capacitor would have made the impedance match more accurate." Canare GS-6 cable has a specified nominal capacitace of 160 pf/m. For a 15 foot cable that works out to be roughly 732 pf. A typical tone control on a guitar can introduce a capacitive load in the .047 microfarad range (depending on the pickup/controls configuration). That's about 64 times the cable's capacitance which means turning your tone control a small fraction of a turn will be equivalent to the effect of cable load. I don't think a switchable cable-emulation capacitor would do diddly squat.
So I think what I've gathered from this article is that first, the author isn't exactly an expert, and that the Zbox may be a good tone changer for those who like its sound, but it definitely does not qualify as a DI. My advice to the OP remains the same. Use the guitar input on your 4Pre...unless, of course, you like the tone of a Zbox.
Phil