That task needed a good number of screenshots a minute. I can get each extraction done with a single keyboard shortcut and selection on a laptop, versus... that a few dozen times. It works okay for a one-off, but there's value in enabling quick non-OS-native solutions to make work easier.
The situation should dictate the solution as decided by the user, instead of the OS dictating a solution and limiting options regardless of the situation.
You can just batch the screenshots and then switch to the photos app for the extraction. I’m still failing to see the problem.
The iPad can use apps in multiple floating windows, there’s no difference here. You don’t have to switch back and forth between apps.
You can even use a keyboard and mouse and the same copy paste keyboard shortcuts if you really want.
And I’m sure there are apps on the App Store to help this workflow in some way. The iPad isn’t some static device where you’re stuck with the software it comes with.
Theres a CLI tool for this gap. textsnap on github runs OCR locally/offline, with no arg it reads an image from clipboard and copies the extracted text back to it. Might not solve the App Store install restriction, but could cut the friction if you can get it running via something like iSH or a Linux vm
Take the screenshot > hit the checkbox > save to photos > open photos app > choose the first photo > select the text.
I guess it’s true that there is a lack of overlapping windows but other than that this process is very quick with very few button presses.