I have tested this method thoroughly, and it seems to be doing the trick quite well. I can't find any errors with it so far, and the physics behaviour seems to be accurate. Thanks for that.
I still stand by my opinion of adding the "set angle of gravity" function to the Physics beahviour of the Construct 2 engine. It's just another improvement to the program.
As far as this thread goes, however, I think this issue can be closed as AnD4D has provided a functional and elegant solution.