My understanding is that pre exposure is more of a technique to support shadow detail. Typically, you would expose (some folks use a styrofoam cup over the lens) at about Zone I or II. The logic is that it makes the most difference to the shadow areas and does not affect the highlights too much. The basic logic is exactly similar to pre flashing printing paper to hold highlight detail - you sort of bring the emulsion up to the speed point. I've never done color using this technique, so I fear my experience might not help too much, but here goes...

4.5(Zone III) 9(IV) 18(middle gray) 36(VI) 72(VII)

The basic logic is that if you add about 2 units of light, your Zone III now shifts to about (6.5) Zone III1/2, Zone IV will be about (11) Zone IV1/4, while Zone VII will show very little shift. Importantly, Zone II (which was originally putting 2.25 units of light on film) will be pretty much at Zone III now (since it's putting 2.25 + 2=4.25 units of light on film).

A couple of concerns with transparency film: Highlights are typically quite critical with transparency film. And you are adding some additional light to the highlight end - in other words, if you've placed a highlight at the edge of film, you run the risk of losing it. I would give myself an extra margin of safety on the highlight end with this technique. Also, transparency film has a pretty steep characteristic curve - so I'm guessing you might see larger tonal shifts at the lower end.

If you wanted to hold a longer subject reflectance range, try rating your film a little slower and then pulling your first development. So rating Velvia at 40 or slower and adjusting development appropriately might help. Alternatively, switch to a less contrasty transparency film like Astia or switch to negative film. There's probably a lower bound to how much you can pull the film (thanks to color shifts), but I would guess that long before that, you'll hit unacceptable loss of local contrast.

If you want to find out exactly how many stops of additional reflectance you can handle, I'm guessing you will need to work with the characteristic curve of the film. Look at the log exposure value for the normal processing curve near the toe and shoulder of the film and calculate the difference (each 0.3 corresponds to a stop). Now look at the similar points on the pulled curve (i.e., the one rated at 40 and given N-1/3 processing). A look at the toe and shoulder log exposure difference should tell you how much you can hold.

Hope this helps. DJ