I completely agree that's what the article says.
So, those that managed to get one of the 200 places, then go on to impress during the internship, and then decide they want to apply for a job, will be 'fast-tracked to the final stages of the Fast Stream selection process'. Given that they've almost certainly had to do the aptitude tests and initial interviews to get a place on the internship scheme, it would be a bit daft to put them back at the start and make them do all that again.
So, some of the 200, will be considered for the 1000 jobs, and will still be competing against the best of the other applicants.
I don't agree that it reduces the pool - other people won't not apply in the first place because maybe a few dozen out of 1000 jobs will go to people on this scheme (it *could* be up to 200, but that seems very unlikely). Given the publicity, and the greater access for applicants from poorer families, then it wouldn't be unreasonable to expect more of them to apply for this scheme, and the main application process, than currently apply. I'd argue that this increases the pool, rather than decreases it.
People from the "chosen subset" at the moment have a lower chance of success, so giving them an advantage in the earlier stages of the application process, doesn't make the overall quality of applicants worse.
I don't know how privileged your own life has been Brewster, but I honestly believe that this kind of process is important, and is one of the main reasons why the civil service isn't still selecting from a small pool of middle and upper class men, as was the case for many, many years. I obviously can't speak for you, but I'd hope you would agree that it's a good thing we've moved on from those times?
I posted later in this thread about some of the experience I had working with people who wanted to be barristers, and I can tell you for certain, that people who don't already have that middle class upbringing, and who don't know people doing similar jobs already, can get completely lost, rather than living up to their potential.
As for you last point - agreed, it sounds nonsensical, but assuming you're going to try something like this, you have to start somewhere. In fact some of those very clever people who previously got through the fast stream system likely wrote a huge report about why it's the best method. I don't know, but it's certainly better than asking Charlie Mullins :)