Yeah, but remember there are far more grad students obtaining PhD's than faculty positions - in other words, sure, not many who make it to faculty leave, but far more people leave than stay overall.
1) Scientist jobs are high-status, you're right, and permanent positions usually come with a comfortable (though rarely really flush) salary and other upper-middle class perks.
However:
2) The number of available permanent jobs as a scientist has been trending down for reasons that have nothing to do with the number of people willing to do those jobs. Basically, research funding has been dropping like a rock.
3) The "professor -> undergrads -> grad-students -> post-docs -> new professor" career model is structurally senseless. There's nothing wrong with having a severe filter on the number of undergrads who become grad-students (in the sense that there have traditionally been many fields where most of the undergrads can get a job that applies their undergrad-level education), but you simply can't set up the entire science career on the basis of an exponential increase in workers without an exponential increase in jobs. The system only even works for computer-scientist types because most of our PhD grads leave academia by default to get applied research or high-level development jobs in industry.
Now, I'm not going to be a pure-academia wanker and say, "Nobody should get a job in industry EVAR!", but neither is it all right to say, "Everyone should just get a job in industry after grad-school!". Companies simply don't want that many PhDs in most fields.
Also, the oversupply of PhD labor and overspecialization thereof in most fields has enabled many universities to shift to a teaching model in which 2/3 of teaching staff are adjuncts without livable salaries, fringe benefits, or any contract for permanent employment.
As much as I really, really like what academia and research are supposed to stand for (and I say this as I current graduate student), I can't give any kind of endorsement to the career model involved. Not while I'm earning approximately $1400/month after taxes as an "entry-level academic" (ie: grad-student) and a friend of mine (whom I decline to name here) with a fairly close skill-level to me earns something like five times that (again, after taxes) working for a major Silicon Valley company.
Hell, I earned three times that much money interning at a Silicon Valley company over the summer! I wasn't even entry-level in industry, and yet I made more than I do in academia and had a clearer, more secure career path in front of me.
The thing I'm thankful for is that we're a bit less "ideological" about the industry-versus-academia split here in Israel. Professorial jobs have never been very well-paid here, so alternating between industry and academia throughout one's career is just considered normal. Oh, and of course the fact that my health insurance and pension funds aren't tied to just one job :-p.