Mutt,
Short answer no. Like most certification performance data, I view Vmcg as a reference or information limit in that, if the real world boundary conditions are much the same as the certification assumptions, then the numbers will more or less line up - real world with the book. This is not to suggest that one may ignore the application of the basic certification rules. Rather, one has a responsibility to consider whether the use of additional fat (margins of safety, fudge factors - whatever sort of term you fancy) may be appropriate when the real world varies from the contrived certification boundary conditions.
Some people probably think that this subject is a pet bee in the bonnet thing with me. However, the facts remain
(a) that a Vmcg departure is very speed sensitive (I wasn't convinced of this until I became involved in a series of flight tests revolving around this matter many years ago).
(b) crosswind adds a not insignificant stabilising/destabilising yawing moment into the cauldron depending on the side of the engine rundown.
(c) simulators may not model the test point with a high degree of fidelity - certainly that has been my experience within a range of sims with which I have worked in training.
(d) many line pilots will, without any additional thought, reduce V1 into the Vmcg regime when, perhaps, there are more appropriate options.
It is up to the operator or pilot, circumstances permitting, to consider adjusting the min V1 on the day to address such real world considerations.
m&v,
The UK CAA required Vmcg to be reduced to the 7kt crosswind case. The has seen the rather silly situation where performance numbers differ for the nominally same aircraft depending on whether the certification basis is US or UK.
The only documented Vmcg variation I have seen having some authority behind it is for the DC9, where Douglas provided some test data indicating, for the particular model with which we were concerned at the time, that the Vmcg increase was around half the crosswind. I suspect that the figure will be applicable with some generality. Certainly, one sees it bandied about from time to time.
[ 08 September 2001: Message edited by: john_tullamarine ]