You gotta credit Amazon with thinking outside the box.
Sounds almost exactly like a project that my team did in my MBA program 2 years ago ... and yes, there were Amazon employees in the class. The benefits of a program like this are trivial in regards to cost savings for retail clerks ... you still need people in the store. Where the big savings come in are in reduction of shrinkage (shoplifting), and reduction in warehouse costs, especially when integrated with online shopping. In short, you no longer need to treat online separate from retail ... every store can be a mini-warehouse and you can fulfill online orders from the stores. When we did our cost analysis, there was about $40M in launch costs, then about $1M per store to implement. Our model was to charge out on a subscription service to the retailers, and it modeled out as a VERY lucrative business.