Suppose that you are writing an email. You mistype someone's address in the "To" field. For example, maybe you write, "firstname.lastname@example.org." Before sending the email, you remember that that person spells their name in a funny way. You want to delete the letter "H" in "john." The correct email address is "email@example.com"
In a significant number of present-day email clients you cannot edit an email address already entered into the "To" field. Once you press enter or shift the keyboard focus to the body of the email, the "To"-address changes. The user interface changes what you have typed into something which can be deleted/removed, but not edited. This is also true of the carbon-copy fields. If you mistype someone's email, you have to delete the address completely, and re-type it. For many email clients, left-clicking on the email address might delete it, but it does not enable text editing.
I am talking about the front-end of the interface, not what's under the hood. What is weird is that in the years from 2000 to 2010, editing a destination email address was trivial. In most email clients, the "To" field was a text-box. You could click anywhere inside of the "To" field and type almost anything you liked. The backspace key worked fine; the delete key worked fine; anything.
Features do not become popular in multiple competing companies user-interfaces, unless those features are an improvement over the old way of doing things. There must be a rationale for disallowing users from editing email addresses previously typed into the "to" field. Technically, you can edit them by deleting and re-typing from scratch, but hopefully my meaning is apparent. What is the thinking behind this? Is it a matter of making "the common case" fast? Which user-cases are faster/easier using the (new or delete) style of design instead of the old (edit text) style of design?