There's a difference between the level of programming required to build a proof of concept, and the level required to make it do anything off the beaten path.

There's a difference between the level of programming required to build a proof of concept, and the level required to make it do anything off the beaten path. I think it's well established that it's never been easier to get a web application off the ground, and UX+design in this space definitely is a differentiator, and probably more important for those initial first impressions than the programming. It's a function of market saturation though, creating your average web application or site is a pretty established problem, with a huge number of options when you need to go about building it. Where engineers are not a commodity? Any company that needs things to be delivered yesterday, in a space where you need cross-disciplinary, out-of-the-box thinking, with the ability to play multiple roles (sysadmin, support, customer service, programmer). These types of engineers are quite hard to find, and satisfaction with the average "commoditized" engineer will be extremely low. Let's say, making an HTML5 iPad app talk to a proprietary hardware device using a proprietary application protocol over TCP to make for a rocking walking salesman appliance. Yeah, good luck making that sing in a sufficient timeframe if you don't know what you're doing.