I would argue that computer scientists (academia) is the biggest offenders when it comes to writing "non-clean" code. They normally don't have any incentive to create reusable and maintainable code.
You might be right. I'm betting on the fact that if you compose your system of small decoupled components you can change your mind at a later time. Lets see what happens :)
I've started working on something along those lines: http://gilliam.github.io/ . Currently using LXC, but adapting to docker is quite easy since "VMs" are provisioned and launched via shell scripts.
I think that he just describes the hybrid between a developer and a product designer. A mix that isn't always good: you shouldn't implement what you design. But for startups it's often the case.
Regarding @geebee "putting developers into a box where they are simply executing specs" I think that the "developer driven culture" has gotten us to the point where we have forgotten that there need to be _a single_ person in charge of designing the product. Otherwise it will be impossible to keep the product integrity. Read The Mythical Man-Month.
Also, the "spec" isn't some kind of pseudo-code. It can be the user manual, a design document or a prototype. Actually, it MUST NOT include any implementation details, since that is up to the developer to decide. It's his domain, he knows that better than anyone else. If the developer feel that he misses out on something, maybe he should try to be a product designer instead. But for the sake of the product, he shouldn't be both.
That's why you write a C-extension to CPython when you need the performance. And if you don't need the performance you stay sane by not having to write stuff in C++.
I'm trying to figure out why people think real-time data is important. Data needs to be actionable, and I'm not sure real-time page view data is that.
What conclusions can be made from that dashboard, except "we have N visitors right now"? And what actions can you take from that data to increase the number of page views?