Not only Orca can read the screen, it can also in case of properly coded applications describe the window (this is the case e.g. of PCManfm), and also spell what you type.
Without it, software only available in graphical mode like e.g. Firefox or Thunderbird would be inaccessible to blind people. Other voices can be chosen and downloaded that could give better results. Bear in mind that providing a good voice is a lot of work. Big players like Apple, Google or Microsoft can afford to pay people who do it, having volunteers to work for months on that is another story.
Indeed it needs some time to get used to it, as with other software, and there is not much incentive for a sighted person to devote the time needed to learn. Unless this person be a packager or a distribution maintainer intending to provide it, of course
