On one hand, I completely agree with Jonathan.
I started my career 20 years ago in games dev and digital advertisement which were strangely close to each other at that time. When I work with developers today who need an external library to check if a number is odd it's terribly frustrating. There are certain techniques like working with binary formats, that were taken for granted even at junior level when I started that seems like a rocket science these days.
When I share my gripes with fellow managers they often joke if I would be a developer on their team and came up with one of my "optimal" ideas they would fire me on the spot, because such solution would be too convoluted and obscure for todays standards and not maintainable.
On the other hand I completely understand that the software industry shifted a lot since I started and now every bussiness is an IT company. This creates a lot of demand on the market and someone has to push these deliverables somehow. And despite what business stake holders will tell you, world won't collapse if these applications cracks in few spots if you'll be able to ship them before competition. People will still continue to use LinkedIn even if it takes 7 seconds to gain interactive on my phone.
Jonathan comes from a different background and his views may seem radical for someone working in an ordinary IT company, but even the game dev industry shifted in that direction. Back when games were released on CDs and you couldn't assume that every user will have a constant access to the internet it was absolutely crucial to make sure that the game will work on day one. But that's no longer the case. Companies like CD Project proved that you can ship a broken build on day one and keep releasing patches months and months after premiere.
I mean, it's embarrassing that GTA V required 5 minutes to parse a simple JSON file for 7 years, but the game sold in 170 million copies regardless.