The part I had a problem wrapping my brain around was figuring in opportunity costs. If you have outside places to put money to earn higher rates of return (like paying down debt), electing for no discount could be optimal. I think the key is to recognize the payment structure has diminishing returns: the one year lump sum earns 10 percent, and the second earns you an additional 5 percent. When put succinctly like that it makes a lot of sense.
I haven't done the math, but for the moment, it seems like if your other investment options are in the 5-10 percent range that guy was right.
EDIT: I built a spreadsheet to verify this. It seems Joe is right, but only if your investment options are above ~12 percent. Under lower conditions, the 2 year deal is better. And above 20 percent and monthly option is better. An interesting paradox of behavioral finance...