I have nothing against todo apps, on the contrary! In fact any improvement in that area/tools can easily affect thousands/millions of people, because organising work/projects/collaboration is needed everywhere.

What I would wish for was that to develop things that don't require native functionality (e.g. only need storage/ram/connection) that there was a unified, consistent, performant, API that allowed one to program against that, and have the owners of the OS provide that API, in this case the browser would be the closest (and many issues people have with it against native apps are solvable).

While it's almost sure that the money and time spent on providing native versions wasn't the sole reason, it just adds a gigantic overhead in many cases - as mentioned in the tweet, it's either time you have your devs dedicate to it instead of all other things, it's the continuous maintenance costs as now instead of having a source of bugs you have N, and new features have to be implemented across N, it's the limitations that it imposes in how and what can be implemented (it has to basically work to the lowest common denominator), the hoops you might need to jump to bring it on pair, and then if you use "wrappers" many times their performance is bad. And if you're using a tool to simplify your work you don't want to be wasting time on every interaction you have with it.

Or you can hire a team to do it for you, which brings costs and still management overhead and dependency or you can choose to develop only for web, but this then comes with the issues of not being normalised well enough.

Have an WebGL interface that scrolls a container and needs to interact with mouse/motion events? Well, outside of Chrome it's going to suck. And etc...

As a consequence, given two ideas that are similar, those with much bigger pockets can outrun the UX and quality of their "adversary" products easily. In the process wasting thousands of hours and $ on just replicating the same thing across platforms.

This was my main point!