Slight catch. if you suspend a rope pendulum from a horizontal string, if the string flexes at all, it will essentially act like a longer pendulum perpendicular to the string (where the string moves) than it does along the string. The result is that the pendulum's swing has a different period along either axis, and it will trace out a lissajous pattern, not a straight line in its swing. And that sort of wrecks the idea of a focault pendulum.
Because I-beams are perfectly rigid and don't flex at all?? I'm guessing "would render that problem so small as to be unnoticeable" but maybe the pendulum is so light that effects are drowned by random noise or don't exist - there must be a potential well to climb out of before lateral forces cause axial flexing so I'm probably wrong.
Oops!
Better make sure your cable is REALLY tight.