Nah, it's really a question of complexity regarding the sheer amount of hardware that needs supported. Apple tightly controls what hardware they write Mac OS for; this lets them make the OS without having to worry about an external ecosystem.

Windows has vendor support by the sheer mass of install base; vendors write their own drivers for Windows.

If you created a new OS tomorrow, where do the drivers come from? If you support too little, then nobody will use the OS.

On top of all that is application support. You need a reason for anyone to want to switch to a different OS, and application developers already need to be there.

Linux (and the unrelated BSDs) has several decades of volunteers writing the apps, subsystems, and drivers based on their own interests. This is one of the reason it works so well in the places it's most used: someone's already done the work for themselves and gave it to us for our use too.

An OS startup has none of those.

/I pine for OS/2 //and BeOS

They can make for an interesting hobby for a certain type of person, but as a startup, it's a long hard road that will almost certainty fail to gain any traction.

If someone really wants to make an OS, they need to make the hardware that OS will run on. It's the hardware that will then sell the device. Releasing a general purpose OS to run on existing hardware is a tough sell.

Pentad.ai — operating system for agents which will outnumber human computer users soon. See pentad.ai/platform for tech details.

How are you going to compete against free and good (Linux)?