It can still be rational, but you need to look at shopping as an aggregate of probabilities.
Let's make it more extreme. A $50 item for free or $50 off a $5000 purchase. Discounts as big as 100% off don't come every day. Maybe you'll never see that deal again in your life. You should jump on it. Where as a 1% discount is something you see every day at dozens of stores. There will probably be a bigger discount next week, so don't bother going out of your way.
If you have to make a choice between the two, you choose the item with the less-common discount.
You're thinking that the question is "should I this item or not", or "should I buy this item now or wait a while". Then it could be logical to prefer the higher percent discount.
But that's not what the study asked. The study said "you need to buy this item now, where will you buy it?" In this case it is illogical to prefer the higher percent discount, dollar value is the only thing that matters.
Let's make it more extreme. A $50 item for free or $50 off a $5000 purchase. Discounts as big as 100% off don't come every day. Maybe you'll never see that deal again in your life. You should jump on it. Where as a 1% discount is something you see every day at dozens of stores. There will probably be a bigger discount next week, so don't bother going out of your way.
If you have to make a choice between the two, you choose the item with the less-common discount.