We used to have bad products that did a hundred different things badly, rushed out to meet an overambitious spec sheet and an unrealistic launch date. Now we have bad products that do one thing badly, rushed out because the Kickstarter money or seed funding has all been spent.
I think the author completely misses the mark on what these teams are getting wrong, and the solution to the problem. The products he describes have clearly had no small amount of effort put into the "whole gestalt experience"; The hardware he complains about looks beautiful, and I expect came in beautiful packaging and was sold with beautiful marketing. His gripes are straightforward engineering issues, mainly of poor software quality. There's nothing romantic about refactoring and debugging. It doesn't take any sort of deep philosophical insight to do proper QA.
If I may be somewhat blunt, I think that the answer is perfectly straightforward - stop buying unproven hardware just because it looks cool. Companies ship buggy, half-finished devices because the market makes it profitable to do so. Curb your enthusiasm and wait for version two. Pass over the hot new thing from the hot new startup, and buy something more pedestrian from a manufacturer with a proven track record. Ignore the glowing press reviews on launch day, and wait for the trickle of complaints on Amazon when the thing starts failing after three months.