Are you guys reading the article? Iron Lore went backrupt (and I suspect the same for Free Radical and Factor 5) because they are too small to do multiple projects and because games requires a large development team now, just the short period of time between going Gold and sales revenue from the game coming in (a couple to few months) is enough to bankrupt them.

It has nothing to do with the game's quality or even sales. Game development has become too expensive.

I hope they pull through even though I don't care for PC or xbox 360 gaming. I just don't like seeing game devs go under.

Even if it was only 1000 copies every day, it would still mean 365k sales at the end of a year. Add the fact that on digital distribution the game will never lose shelf space and will keep selling "forever".... it's a win.

I already mentioned this before (I'm not sure if it was to you), the only figure we know about Steam is that Garry's Mod sold 312k in 2 years. It was released in 29th November 2006, and sold 5.7k in the first day, then went on to sell an average of 430 copies per day.
Now keep in mind that today Steam is several times bigger than it was in 2006 and wasn't as good as today... it was only after 11 months of Garry's Mod launch that there was a "BOOM" on Steam with Orange Box (Oct. 2007) and also with L4D/FM2009 (Nov. 2008), and the game hasn't charted in the last few months of that 2 year span from what I remember.

Steam is growing mad as hell. Here are some very rough estimates to give a more general view:

Nov 2006: Garry's Mod sells 5.7k in the first day, with 430 per day for the next 2 years => 312k.

Steam increases over 150% in 2007. Let's multiply the games' sales 2.5 times:
Nov 2007: "B" game sells 14.3k in first day, with 1075 copies per day for the next 2 years => 784k.

Let's say Steam increases over 150% again in 2008 (very probable), and multiply the games' sales 2.5 times:
Nov 2008: "C" game sells 35.8k first day, and 2687 copies per day for next 2 years => 1960k.

These are very rough calculations though, and many of the Steam sales increase is due in part of the staying power of the older games, so new games' sales aren't really increased 2.5 times. Also, there are much more new games coming out than there were before.

The upside is that Steam games will still keep selling after 2 years.

I think the real problem is that Western developers spend to much money on graphics because they think graphic whores in the majority when they are actually the minority.  If they spend less money on graphics and just make a game even if the graphics aren't groundbreaking, this would happen less often.  Also, this is bad because small developers tend to have more creative control over games than big publishers.  This means that we will get a ton of FPSs and most of them will suck.

What on earth are you doing here?

You can't apply an increase in total sales on Steam on individual titles like that. You're just guessing.

The Steam sales increase comes most likely from an increased number of games being released on Steam and their back catalog getting bigger, not just from more people signing up.

It could well be the case that individual titles actually are getting lower average sales as more and more games are being available on Steam.
(since much like gaming consoles early in a gen, you had a lot of people signing up on Steam because of Valve's games, but they didn't necessarily have a lot of games to choose from - thus attach rate goes down drastically over time)

Smaller developers tend to be more efficient, money-wise. Just look at Ironclad, they made Sins of a Solar Empire with a team of 4 guys and a budget of less than $1 million. In my opinion Iron Lore just had too many people working for what they were, and closed down before they could enjoy Soulstorm's sales.

Dude, that's what I said in my post you quoted. I bolded it for you.

Anyway, I couldn't find how many games there were in earlier times, so I can't know. But either way, more and more money is being spent on Steam, and there must be a reason for even EA joining Valve.