I do not have a definitive answer, but I do know that calculating the V-speeds is more complicated than just drawing a graph and identifying a specific point on the curve.
At the end of the day, for best rate of climb (Vy) or best angle of climb (Vx) calculations, you've got to superimpose three graphs on top of each other. You've got to start with the airframe efficiency graph (L/D at various airspeeds), and then superimpose engine efficiency (torque/HP at various RPMs) and propellor efficiency (net thrust at different airspeeds and RPMs). And of course things may be complicated further because at high airframe AoA the thrust vector is not in line with the flight path (drag vector), the whirlwind effect of the prop around the fuselage may create additional drag, the touch of right rudder required induces drag, and so forth.
My gut feeling is that the Oxford text and videos will only take the airframe efficiency graph into account, and assume that thrust (or drag) is a constant. They use that simplified model to argue that Vy = Vmd. And at that stage of training, that may well be the proper approach, especially if the students will eventually fly turbine driven aircraft, who have totally different characteristics than piston aircraft when it comes to power delivery at various airspeeds and altitudes.
In real life, I would assume that Vx and Vy are not calculated, but measured by the test pilots during the test program. And these real-world values will automatically take all inefficiencies into account. I can even imagine that test pilots / POH writers will fudge the numbers a bit, for instance if the real-world Vx number would only be achieved at airspeeds that would not allow for sufficient engine cooling.
In theory, there is no difference between theory and practice. In practice however, there is.