I must admit I hadn't thought about that, but I don't think it is the cause of the problem. You have taken a distance that the pendulum moves and divided it by the time taken. That gives you a speed, but not the maximum speed. Your pendulum starts at rest and accelerates, so it is moving faster and faster as it falls. The speed is varying, but your calculation gives only a single speed - the speed it would need to go at if it moved at constant speed, ie the average speed. (Actually the average speed if it just fell vertically, but changing the distance to an arc would still leave you with an average speed.)
I'm not sure what the question is getting at, as I'd just use the energy equation to work out the ideal max speed. But in that case you don't need the information about the camera. So I wonder if they want you to use that information to estimate frictional or air resistance effects? I haven't done that myself, so I'm not sure off hand what you could get. You know only the total time, the geometry, the position at start & end, gravity and the initial speed. What sort of resistance function you could estimate from that, I don't know.
Your own question about the discrepancy is simply the difference between the instantaneous speed at the bottom of the arc and the average speed over the arc. (Your 4.738 would be even smaller if you calculated correctly, but that's irrelevant.)
I'm going qrt now, but I'll see what I can work out tomorrow, if you're still not sorted.