.. a few things to note
(a) Vmcg is a line in the sand certification animal for nominally critical conditions. Like everything else to do with certification, it is not a guarantee, only a benchmark considered to provide an adequate risk level.
(b) usually, the figure is not particularly relevant to real world operations due to the configuration's being different (especially CG and actual thrust) .. this has the result of lowering the real Vmcg on the day .. however, the certification figure still determines the minV1 boundary
(c) based on the aircraft for which I have had a Vmcg test involvement, the handling changes in the Vmcg region are very rapid and quite critical. That is to say, a continued failure just above the Vef for Vmcg will be OK and the aircraft will deviate and then return to the centreline. In the space of a few knots reduction through Vmcg, the deviation will rapidly become uncontrollable and a close encounter of the grassy kind becomes the predictable outcome if the throttles are not closed very promptly
(d) generally the continued takeoff is the critical case, the accel-stop being comparatively straightforward .. subject to an appropriate pilot response
(e) when the NWS physically can be disconnected for the tests, that is a usual test configuration
(f) with max aft CG, the nosewheel load is going to be very light and, if the event occurs for critical conditions, by the time the average startle-affected pilot gets the pole forward .. it is probably a case of too little, far too late .. For instance, I had a period training in a 732 sim which appeared to be reasonably realistic for minV1 failures. The particular operator appeared never to address this case as they routinely used significant overspeed takeoff schedules .. it was interesting to watch the open eyes on the crews when I started exposing some of them to the problems ...
(g) the centreline deviation varies with the certification basis but the philosophy is the same regardless of the maximum deviation permitted
(h) generally Vmcg is determined for nil wind. The effect of crosswind in real world considerations is critical and may increase the real world Vmcg above the certification animal in the order of 0.5 kt/kt (twins) to in excess of 1.0 kt/kt for four motor aeroplanes ... an important consideration for the selection of speed schedules (or considering delaying departure) if such options are available
(i) for some aircraft Vmcg is so low that it is not limiting (eg Citation)