The inherent problem is ad money spent by retail locations don't convert to more foot traffic like a Google ad can send more eyeballs. And incentivizing people already at your store with coupons or via an iPhone app is awkward.
FourSquare is the 21st equivalent of billboards, lots of voodoo math and promises involved. And some people get a cheap thrill out of tagging up boards but in the end it's still just a tagged up billboard out on a lonely highway.
But, hey, who would be dumb enough to turn down $20 million and a shot?
It is far too common to think that less measurable = always bad. It's just not true.
I've never used Foursquare advertising and have no idea if it works, but there are a lot of unfounded "conclusions" here. Think about a restaurant that makes $20/customer, servers 200 customers/day and is open 30 days/month. That's 120k/month. Let's say the 4-digit campaign was $5k and had a 150% ROI, so it generated $7.5k in business. That's 6.25% of the monthly revenue, an amount I could certainly see being hard to parse out due to normal variations.