What makes the experiment so interesting is the anonymity factor. Since there are a functionally-unlimited number of Starbucks locations, you have absolutely no way of knowing who else is using the card and on what.

If it was for a small locally-owned coffee shop, it's a given that the folks charging/refilling the card are the same people frequenting the shop with you every day. I can't imagine this wouldn't drastically affect the way the experiment turned out; the social obligation is much stronger when you can connect anonymous people to faces or names, even if you don't know exactly who is participating.