You can clearly see once you begin your aileron rolls, you perform them in the same axis with barely any meters of displacement (docs show a single roll ends with at least 100-300 meters)
For reference an accurate High G Barrel Maneuver :
Maneuverable counteraction to missiles EN.pdf
Doesn't exist, its a maneuver born out of a simulator that has yet to fully develop an accurate missile guidance system. To pretend that its a real thing that can be employed in IRL as I said is naive at best.
Not providing tacviews or tracks doesn't prove your point at all and in fact makes it look worse at your attempts to confirm that "is a real maneuver" requires reproduction and study.
And there is no point in studying the missile behaivour in DCS since as ED said they have to fully finish an accurate guidance model, maneuvers IRL can apply to DCS but not the other way around.
Whether this happens on how the target velocity vector information is given to the missile, delay in ms of both internal missile logic + simulated hardware hickups or simulated frame time or missing documentations of software algorithms for late stage missile systems is for ED to deal with and study.
This maneuver has always been an exploit, nobody cares if it's an AoA desync roll that causes teleporting or "resonant roll" as you call it, there is no evidence that a maneuver like that would work IRL and as far as feedback comes from, examples and documents from reality are the go to in order to achieve a believable state not things born out of DCS, it would be like training AI datasets on AI hallucinations.