I did say I was guessing didn't I, but...
(1) The strict definition of lateral stability is rolling moment due to sideslip.
(2) The strict definition of directional stability is yawing moment due to sideslip.
Just checking the sign of lateral stability, we know that a normal aeroplane can be turned on the rudder, if a little innefficiently. So, put in right boot, rudder deflects to the right. This pulls the tail left, and the nose right, creates sideslip from the left, and we know that the aircraft will therefore roll to the right. Damn, there goes my logic, lets try again - I've just looked it up and it seems there's a mild conflict - what as an FTE I know as positive static lateral stability actually has a negative sign (Mike Cook's FLIGHT DYNAMICS - PRINCIPLES is my reference of the moment, not my favourite text on the subject but its available). I think this goes some way to explaining my incorrect explanation, what to us normal folks is positive lateral stability, to an aerodynamicist (I were one once but grew out of it) is actually negative.
So let's look at this again:-
I am starting from the assumption that SP is telling us the truth, and he really does fly Cessna singles. So, working backwards:-
(i) Aircraft rolls left without aileron input
(ii) Assuming that lateral static stability (Lv or L-sub-Beta depending upon whether you prefer UK or US terminology) is a player, and that it is in the normal sense (+Ve to a FTE, -Ve to an aerodynamicist) (I'm pretty confident on both scores so far) then there must be sideslip from the left, that is a yaw string would be deflected to the right.
(iii) So something has happened which causes the nose to go right (or the tail to go left). We know that the only thing that's been done is the right door has been opened.
(iv) Now the last time I swallowed my pride and flew a Cessna I seem to recall that the door was hinged at the front edge, so opening it will effectively create a bulge on that side.
(v) Now conjecting moderately wildly, let's say this effectively creates the top surface of an aerofoil and pulls the nose to the right when the right door is opened.
(
vb) Alternatively, opening the door creates disturbed airflow down the right hand side of the aircraft, thus lower airspeed flow over the right hand side of the tail, causing the tail to tend to pull to the left where the flow is smoother.
(vi) Which completes the proposed mechanism. Pretty much the complete opposite of what I said first time, but what the heck - at least this model seems to fit the data.
I hope somebody follows this, 'cos I'm not particularly convinced that I do.
G