My reason for answering 1A is that I think instamoving the paddle is "cheating", and the mouselag solution is just waay too frustrating.
IMHO this is part of why game design is so hard. Question 1 is almost enitrely useless for the designer - because everyone who reads it will re-interpret "slowly" differently.
Having played lots of Pong, I'd go for the drifting paddle, on the assumption that you will start it slow, find it sucks as a game, speed it up a bit, and eventually get it to follow the mouse at a speed that works nicely.
If you want to test people's choices for *analogue* controls like this, you now need to prototype. Just write an applet in 15-minutes that lets you move a paddle left and right with different control mechanisms (hit a, b, c to change) and different speeds (hit 1,2,3,4,5 to change - for option b the speed can be ignored!)
You may find that people change their mind when exposed to different speeds (including me - I wouldn't rule out that I might find option c irritating at *every* speed, although I'd be a little surprised).