If your goal for something like this is PR, you want a big first-day number you can give press to try and gin up a second round of stories about the campaign. So it can be worth it to potentially lose a little on those sales if by doing so you can get more of them into that first-day figure.
Remember, Mark specifically says that most manufacturers won't take you seriously unless you order 100,000 units. At $600 per phone, that's only 53k units (the volumes look worse at $830 per phone). They can't actually manufacture the phone as described for $32M.
So, this is, IMHO, a publicity stunt intended to get carriers interested in Ubuntu phone, and it probably was quite successful.
I would wager that they can't even buy RAM directly from Samsung without an order over 100k units.
So yeah, maybe $600 per phone is high, but maybe not. They're talking about a serious bill of materials, and although they're economizing on the CMOS and the LCD for the screen, I don't think that will reduce the cost much. The sapphire screen and the single-cut shell are both significant cost structures.
It's fun to argue about unknowable things, but given that (I assume) neither of us works for Canonical, I think the answer is probably that you're right. It would be hard for that phone to cost $600, but it won't be cheap.
http://hup.hu/cikkek/20130723/indiegogo_kampanyt_inditott_a_...
Rough translation of the first sentence ("Ennek az akciónak elsősorban az a célja, hogy az Ubuntu híre még jobban eljusson az átlagemberekhez."):
The aim of this move is to raise awareness about Ubuntu among average users.
I looked at it, was interested but not sure. Now the $600 is sold out, I feel lees likely to buy one, even though that is partly irrational (I should just evaluate whether it is worth $830 to me, not be upset I missed out on the cheaper one).