Although VLD (secant ogive) bullets have the reputation for needing to be seated into the lands, I find that to be somewhat anecdotal. Certainly the transition of a secant ogive bullet from ogive to bearing surface is far more "abrupt" than a tangent ogive bullet, and it could easily affect the way they enter the lands. In my hands, VLD bullets do tend to be slightly more finicky with regard to seating depth than tangent ogive bullets. Nonetheless, I have developed plenty of loads with VLD bullets seated off the lands, and they shoot just fine if you're willing to rigorously test seating depth in fine (.003") increments. Any more, I start pretty much every load development, regardless of the bullet used, with the bullet seated at .015" off the lands. Once a satisfactory charge weight has been identified, I will start by testing seating depth in .003" increments from .006" off to .024" off. Notice that the initial charge weight seating depth of .015" off the lands in exactly in the center of the total seating depth test range. That is so you're not changing the seating depth more than one half the total range in either direction, which in my hands is not sufficient to dramatically alter pressure or velocity with jumped bullets. I have rarely found a bullet that wouldn't tune in somewhere within that range; however, it is certainly possible.
In the event I don't find an appropriate seating depth between .006" and .024" off, I might try seating bullets next at "touching", .003", .006", and .009" into the lands. I have also found that some VLD bullets will tune in very nicely somewhere in that range, but the optimal seating depth window generally seems to more narrow than for the seating depth optima off the lands.
If you use the Berger VLD bullet seating depth approach, just remember that it is meant as a very coarse test to find one seating depth region where the bullets shoots better than the others. In other words, you're testing a very wide range in coarse increments to narrow down the testing zone. You still want to go back cover that single region using a fine increment test to optimally tune tune the load. Also, changing the seating depth of a bullet that far, both into and off of the lands, can create pressure issues as you may be changing the effective case volume by a significant amount. So it's not a bad idea to use a slightly reduced load for safety reasons. Many people have found Berger's approach to work very well, especially for bullets that did not seem to tune in well using a more standard approach.
Paddyd, FWIW - there have been a few reports here that some of the RDF bullets seemed to prefer a fairly good jump, maybe in the .030" to .050" (or more) range.