While it's quite possible to build an app that way, using scraping. The danger is that users of that app will get suspended from Twitter (users have explicitly agreed not to scrape), once thats happened to a couple of users, thats the end of your app.
Obviously you can try to obscure that you are scraping, but that is difficult to get right and there is no way back once you get it wrong.
On Apple devices that wouldn’t be an issue because you you can load the Twitter Web interface into a real Safari and get the markup from there for processing. Twitter can try things like CAPTCHA but this will degrade the experience for all Twitter users and you can simply delegate this CAPTCHA to your user to solve anyway.
While for reading that works, there is an easy why via posting to spot people use that app. They can change at any moment the input names so that when posting a tweet you are exposing that you are using the scrapping client.
Yes, but the app is not a human and will likely interact with the website in a more predictable (and thus, detectable) way than the human itself would.
Obviously you can try to obscure that you are scraping, but that is difficult to get right and there is no way back once you get it wrong.