At the entrance to a ciurve, or at the transition of a reverse curve pair, you would ideally like the pivot of the 2-wheel truck to be close to the truck axle, but in the curve itself all the way back to the center of the drivers' wheelbase.
I have always looked at this a bit differently: the 'ideal' place to put that pivot would be the center of the rigid wheelbase, with good lateral motion on the first driver pair(s) doing some of the curve-shock attenuation. Where the 'variability' is probably best applied is in the lateral compensation. Stability is a very different thing for a lead truck using Voyce Glaze-style overbalance, as the truck has to control the yaw forces (including any momentum) in addition to its own curve-following and self-guiding tasks.
Perhaps interestingly, a pin-guided truck in the trailing position is nowhere near as dynamically stable as a good Bissel with angled rocker compensation, at least when the locomotive is running forward. Reading discovered this and a couple of other things that should have been clear from good contemporary fabricated trailing-truck practice (and would become even more clear with the second Delta design) when they experimented with symmetrical pin-guided trucks on their 4-4-4s just before WW1. (I do not know whether the two Baltics from just a couple of years earlier had similar issues with their pin-guided trucks -- this by the way is why a 4-6-4 with a Bissel rear truck is a Hudson, but one with a rear pin-guided truck, commuter double-ender or not, is a Baltic, except on Milwaukee). The Germans didn't remember this when designing their high-speed tank engines in the '30s, and at least one of those had an air-activated mechanism that moved the pivot point to give stability in both directions... if I remember correctly, as-built the critical speed was no more than about 81mph so the air device was beneficial.
... complicated geometrical control is difficult, hard-to-maintain, and seldom accurate under all conditions.
The actual geometric accommodation needed for a 2-wheel lead truck to be reasonably stable when provided with appropriate lateral restoring force is not that difficult, and much of the actual curve-following physics is accommodated in the fixed wheelset in the same way it is according to Wickens. Were the two wheels permitted to revolve at different speeds (e.g. on stub axles, as they would be if for example you wanted to implement some kind of Ackerman steering) the situation would become more difficult, and consequences of some comparatively likely system failures (including resonance effects) could quickly destroy any nominal advantage from that idea.
It may also 'pay' to remember that these trucks carry considerable engine weight, and some of the effective weight transfer that occurs in curving will need to be accommodated in the truck suspension and compliance effectively. If this can be done in a way that aids nominal stability, I would think higher speeds than those predicted from static geometry might be achievable.
Was Wilgus connected with the BRT, as well as the Central? (Malbone St. reference)
Not that I know of. It was just a bit of black humor: there was a terrible wreck just a couple of days after the GCT electric service started, which according to Staufer was caused by the same kind of wild overspeed observed in the Malbone St. wreck. Very promptly -- and apparently without involving Mr. Wilgus, the engineer in charge -- the engines were sent to Harmon to be fitted with four-wheel trucks, something Wilgus thought highly unnecessary and a slur on his professional ability.
Staufer's account of all this was written half a century after the events, and it will be interesting to see what Mike comes up with, both in newspaper accounts of the situation (including Wilgus resigning right at the moment of his greatest triumph) and in engineering discussions of vehicle instability.