Like hell are they going to kill off a service that gives them access to private, unique, mineable data about someone's personal and work life in order to crawl it for their interests and market to them. An article written with little or no understanding of how Google makes, and will continue to make, most of its money.

I like this, and think it's a cop out in equal measure!

The key is in the middle ground. What can you do in the time and budget given to you to enhance that core experience and bring the user real utility and furthermore, delight?

I've never quite got into these.

To me, if I'm designing a site and deciding how it will scale, I want to decide on the right way to implement a column system myself, not fit what I'm doing into a framework.

Not saying they're bad, I've just never managed to reconcile the 'how' of many grid frameworks available with the 'why'.

Also in response to Robert, the real question is if with a minimum of technical endeavour you can have your business' brand identity look crisp on high pixel density displays (which are far more prevalent than just Apple devices), how does looking smoother and more professional not fit with startup or business goals?

The reality is that photographs are not a huge issue on retina displays, because the information they contain is 'fuzzy-edged' anyway - most of the time, you'll get away with a standard resolution image.

For brand ID, vector-esque graphics and icons, however, would you really consider looking fuzzy and jagged-edged on a decent proportion of smart / mobile devices (and, slowly, laptop/desktop machines too) beneficial to business? Sure, some visitors won't care - but potentially, the ones that do could be turned off by something that's easy to fix.

The best outcome from this kind of discussion is looking towards ways of beginning to use vector assets on a site - not just using pixel density media queries to switch out bitmap images for retina displays, but using tools like Icon fonts - or better still, SVG with a PNG fallback for devices that don't support SVG (and there's pretty good crossover between the pool of devices with higher res displays and the pool of devices supporting SVG in the browser).

That's why I've been using Filament group's Grunticon, which is a Grunt process that takes a folder of .svg files I give it from Illustrator, and turns them into base64 encoded SVG in one stylesheet, base64 encoded PNG in another, and a folder of individual PNGs with a third stylesheet for browsers that don't support either of the encoded formats. Stick their JS loader in to select the appropriate stylesheet for the capabilities of the browser, and suddenly you have your own single HTTP request, resolution independent way of serving the best quality vector-esque graphics you can to all devices.

Then you can stop caring about whether in future the balance is going to tip and you need to start factoring in these high resolution displays as they become more and more prevalent, because you're already using a future-friendly solution.

For us, Responsive and resolution independent are the baseline experience. Client's don't always know what's best, so we've taken the approach that how we build things is more our concern than theirs - and if they consider any of that 'extra' they can find a bargain basement provider instead.

Correlation doesn't imply causation. The article is correct in drawing out that flat design is becoming something of a trend, but to paint RWD as the reason for this is to miss the fundamental aspect that RWD is a methodology and nothing more, prescribing literally nothing in terms of the aesthetics of a web experience.