Sitting in the airport waiting for my connecting flight, so I've some more time to continue. One key realization which is often forgotten in these types of fora discussions is that in the end it is about there being a market for the game you're trying to make, and that market being sufficiently large to pay for not only the development but also all the costs involved in releasing the game. That sounds straightforward, and I often get the impression when browsing through forums that most gamers realize that, but what they usually don't realize is exactly what kind of numbers are involved. That's not so surprising, given the amount of distorted numbers that appear in the media. It's in vogue of late to claim you sold a gazillion units whereas in reality you only sold a fraction of that and I think a good games journalist might actually have a lot of fun (and hits ;)) if he'd care to investigate what the reality is behind the numbers.
In the best case, they are the sell-in (not sold-through) numbers, in the worst case they are flat out lies. The reality is also that it's pretty hard to get a hold of these numbers, which is what allows for the inflation of numbers in the first place (as they're not publicly available), and whatever industry charts there are only give you but an impression of the real figures, not the actual sales. They're sometimes of up to 50% off (in retail), but of course you do get a relative idea (which is how I know that some people are making some very big claims which aren't matched by reality at all)
Right, I've got to break off again - a couple of calls happened during the writing of this post and now I need to run

I am trying to make a case with all of this, but it's quite elaborate, so I'll need to continue it another day. I wanted to make the bridge by providing you with some sales numbers/costs of running a studio/genre numbers to show you why the entire hardcore aversion exists among publishers, but it'll have to wait.
Cheers from London