A broader question might be, "Exactly what are we doing when we tune seating depth?". I've heard many explanations including optimizing how the bullet enters the rifling, barrel timing/harmonics, how the case neck seals, etc. But the explanations I've come across seem largely anecdotal. The answer to that question, if there even is a well-defined answer, might shed light on why VLDs have a reputation for being finicky with regard to seating depth.
FWIW - my thought is that part of the reason for this reputation may have originated years ago when VLD bullets became more commonly available. I suspect, although it is only my unsupported opinion, that many started using them in commercial rifles, such as hunting rifles, that were manufactured with chambers having an extraordinarily long freebore, such that the VLDs were necessarily loaded with very long jumps. With their rather abrupt transition from ogive to bearing surface, I can imagine how this might have caused VLDs to get a reputation for being jump sensitive that still seems to be commonplace even today.
I have used Berger's 90 VLD in F-TR competitions for a number of years. Initially, I tried seating them at .010" into the lands, as that's where the relatively small amount of information available on their use at that time suggested others had obtained successful results. However, in my hands the groups at -.010" were mediocre at best. So I started moving them incrementally out of the lands in an attempt to find a more optimal seating depth. The end result was that the 90 VLDs appeared to have at least two optimal seating depth windows. The first was with them seated into the lands at ~-.004" to -.007". The second window was between about .018" to .024" off the lands. Both shot equally well, but because the second window was not only wider, but also did not involve jamming the bullets, that's where I chose to seat them. Since that time, I have observed the 90 VLDs to show that exact same pattern of seating depth optimization in different rifles and multiple barrels. So it seems to be reproducible, in my hands at least. Further, although their specific seating depth optima clearly differ from that of other tangent or hybrid ogive bullets I routinely load, I would not call the 90 VLDs any more "sensitive", merely "different".
The bottom line is that the factor(s) we are optimizing via seating depth adjustment may be very, very complex, and therefore difficult to define with any precision. Nonetheless, it is relatively simple exercise to conduct a seating depth test and determine empirically where a given bullet wants to be seated. So although we would very much like to know exactly how/why seating depth has such a profound effect on precision, we can still determine optimal seating depth for a given bullet by testing without necessarily understanding exactly what is happening.