DZ,
I'm not doubting the significance of intake momentum drag, which you quite rightly point out can cause yaw leading to sideslip which might lead to uncontrollable roll, depending on the IAS and AoA. The film to which you refer does indeed show such a sequence of events. On the sound track, you will hear the (American) advice from the ground to accelerate away, but phrased something like "Stow the theta-J lever". IMD in yaw is particularly noticeable in the Harrier as it makes it directionally unstable at low speed - it always wants to point out of wind. In a 360 degree spot turn, little rudder is needed for the first half, but more has to be applied to turn back into wind.
I once saw another pilot (co-incidentally also American) survive a similar incident to the one in the film. He was decelerating with a strong crosswind when he let too much sideslip develop, and he took a while to catch up with the rudder action while trying to accelerate away. This resulted in a rapid yaw through 450 degrees followed by an acceleration to wing-borne flight 90 degrees off his original track. Not only did this scare that pilot, and also quite a few of us spectators, but not least the pilot of another Harrier decelerating in the same original direction, trying to cope with the crosswind, who suddenly got a windscreen full of jet flying across fairly close straight in front of him!
I seem to remember a filmed example of IMD in pitch with an F4. I think that it was a new aircraft on a delivery flight (from St Louis) that just carried on rotating up until the crew ejected.
By the way, thinking of St Louis, I always found it a bit odd when confronted by the huge sign over one of the freeways there proclaiming "St Louis - Home of the Harrier".
Returning to the original subject, there was a fixed relationship between the head-up sideslip ball and the shakers. I think that the point at which the central vertical line became tangential to the ball equated to the 0.06 lateral G at which the appropriate shaker operated.