I personally think this thing will be misused to the hilt, but there are ways they can add real value to a brick and mortar shopping experience.
Imagine you walked into a larger store, for example Walmart. As you walk in, the estimote app loads the wallmart app temporarily over the instore wifi. The app shows you a map of the store, which shows what products can be bought in which aisle, making your shopping experience much faster. As you approach the cash register, it acts like a concierge and tells you which register is the fastest one (some airports do this with a dedicated person, which directs people to different lines. It yields considerable speed ups). As you leave the store, the walmart app is removed from your phone. Estimote now shows you the nearest estimote enabled stores.
That's the good scenario. Now to the bad.
You go to the mall. As you approach it, you phone starts buzzing like it's mad. You think it's a call, so you pick up your phone. Nope, its Footlocker trying to sell you their new line of shoes at a 10% discount. Fine, but you're not looking for shoes. As you walk through the mall, your phone goes of wantonly, displaying ads for stores that are on the other side of the mall. Obviously, they don't know how to configure their estimotes. You can't walk for 5 meters without an estimote alert going off, logging itself to your allready estimote infested task list. Completely pissed off, you turn off your phone in order to leave the mall at peace. You're not buying stuff here. As you get home, you get an important call, but your phone is off, and you miss it. Resignated, you trash your iphone and get yourself a single purpose nokia. Serves you right estimote, serves your right.
I think there's certainly potential behind these, but their "sell the dev packs" approach to it doesn't seem to be very promising. Most small retailers will struggle with the setup and SDK. Hell, the local supermarkets cash registers here have SQL injection vunerabilities. One bar code that includes ;DROP TABLE * in it and they're closed for a week.
There seem to be several steps in setting these up.
1. Buying them.
2. Setting them up physically.
3. ? Linking them to a central system ?
4. Writing an app.
Out of these points, it seems they are only covering 1, even though there's certainly work involved (and money to be made) in steps 2 - 4 . Dumping these on the buyer leads to friction in the buying process which will cost them sales. They need to sort that out.