Here in southwestern Europe most people live in apartment buildings. Their design is usually guided by cost concerns, developed and approved by people who never set foot in them. As a result, most of us end up living in apartments which don't really work. Mine has a niche for the washing machine but nowhere to dry the laundry, for instance.
Reading the op's post, I wonder if there's room (excuse the pun) for hackers to disrupt the house construction market with affordable, highly functional units that make their tenants' lives better in a much more tangible way than another photo sharing app.
The Dymaxion House was developed by inventor and architect Buckminster Fuller to address several perceived shortcomings with existing homebuilding techniques. Fuller designed several versions of the house at different times, but they were all factory manufactured kits, assembled on site, intended to be suitable for any site or environment and to use resources efficiently. One important design consideration was ease of shipment and assembly.
Feels like it needs a conclusion - if the method of designing websites doesn't work well for architecture, then what happens when you try and design a house like a website: what works surprisingly well, what leads to a less-than-optimal house or one that just feels wrong, etc? Does this demonstrate the author's original theory of the architecture flow of work being better, or does the beautiful house designed at the end of the article suggest that different workflows can still lead to a good product?
A diagram demonstrating a series of functional/spatial relationships is the most distant thing from design in architecture. Although the relationships are important and may need to be maintained for the project, they rarely map literally into the projects actual physical embodiment.
Architecture is an utterly holistic process that unifies physical structure, conceptual structure, aesthetics, finishes, functional programming (space planning), user experience into a physical entity that can be deciphered by its users. Designing a house "like a web application" addresses only one tiny part of this process. Maybe the author intended this as a critique of the process of designing a web app (I'm not sure, his conclusion was very brief) being too limited in its scope.
Reading the op's post, I wonder if there's room (excuse the pun) for hackers to disrupt the house construction market with affordable, highly functional units that make their tenants' lives better in a much more tangible way than another photo sharing app.