TL;DR: Open source software has high costs to adoption. The lack of standardization leads to unnecessary work. Proprietary platforms with stable tool chains are more efficient to use. Therefore, everyone should use proprietary platforms.
The author has no problem ceding the entire future of computing to a few mega-companies that have the scale the 'lithify' all the components of a platform. I doubt I am the only person who feels that no amount of efficiency gains would justify giving in to that future.
Most open source software never reaches version 1.0. Few projects gain enough support to compete with polished proprietary software. The answer isn't to abandon open source, but to increase support for existing projects. If no good open source software exists then use whatever proprietary software you need to, but also find the best comparable open source project and contribute to it. Paying for proprietary software but not contributing to open source is like paying interest on debt but never paying down the principal balance. That course of action will never lead to the debt going away.
The other problem with the author's take is that he ignores the other solution to high transactions costs, open standards. There is a reason why we can visit any site on the web with zero transaction costs. The web enabled cross-compatibility by creating open standards for sharing and viewing remote content. Open standards like OpenGL, WebGl, ActivityPub, Open Document Format, and RSS do a lot to allow compatibility between different platforms. Standards like these can allow open source software to match the functionality of proprietary platforms.
Open source may lag behind in a few notable areas. Syncing application files between devices is a pain. Discoverability within federated networks is poor. Quality between applications varies a lot. Open source isn't perfect but the goal of public domain software is non-negotiable. As more economic and social processes become automated, the knowledge consumed within these processes will move from the brains of workers and into the software that runs the machines. This is a massive transfer of power from labor to capital. If this software is owned and controlled by a few large corporations, which would be inevitable if all operating systems were proprietary, then these few corporations would wield significant control over all economic and societal processes. This is as dystopian of a future as I can imagine and this seems like the only result that could come about if we followed the advice in this article.
The author has no problem ceding the entire future of computing to a few mega-companies that have the scale the 'lithify' all the components of a platform. I doubt I am the only person who feels that no amount of efficiency gains would justify giving in to that future.
Most open source software never reaches version 1.0. Few projects gain enough support to compete with polished proprietary software. The answer isn't to abandon open source, but to increase support for existing projects. If no good open source software exists then use whatever proprietary software you need to, but also find the best comparable open source project and contribute to it. Paying for proprietary software but not contributing to open source is like paying interest on debt but never paying down the principal balance. That course of action will never lead to the debt going away.
The other problem with the author's take is that he ignores the other solution to high transactions costs, open standards. There is a reason why we can visit any site on the web with zero transaction costs. The web enabled cross-compatibility by creating open standards for sharing and viewing remote content. Open standards like OpenGL, WebGl, ActivityPub, Open Document Format, and RSS do a lot to allow compatibility between different platforms. Standards like these can allow open source software to match the functionality of proprietary platforms.
Open source may lag behind in a few notable areas. Syncing application files between devices is a pain. Discoverability within federated networks is poor. Quality between applications varies a lot. Open source isn't perfect but the goal of public domain software is non-negotiable. As more economic and social processes become automated, the knowledge consumed within these processes will move from the brains of workers and into the software that runs the machines. This is a massive transfer of power from labor to capital. If this software is owned and controlled by a few large corporations, which would be inevitable if all operating systems were proprietary, then these few corporations would wield significant control over all economic and societal processes. This is as dystopian of a future as I can imagine and this seems like the only result that could come about if we followed the advice in this article.