It's just good business sense, especially for smaller studios that are not backed by a big publisher. They can invest a relatively modest amount of money into a game and expect a realistic return on investment if they go with digital sales.
Just look at Torchlight- started off as 17 people, took them 11 months to make. 2 years later it's still selling and has passed the 1 million units sold mark. Runic more than made their money off of that game and will continue to get some income from it over the next few years (admittidly less) and it'll be bundled with Torchlight II collectors edition or some garbage like that, you know the drill...
All that being said, if I ran an independant game studio this is pretty much what I would be doing as well. Spread out the risk across a few smaller games, focus on making them fun, fast to develop and cheap to buy. If people feel like they are getting a lot of fun for their 5 or 10 dollars, then it's really a no brainer and you'll more than make your bones back.
Also, and I can speak to this because I run a large software department, getting your team experience like this across a lot of smaller projects is very valuable. It helps smooth out the work process and gives the opportunity for a lot of skills to develop - techniques can be tried out in different ways across projects and then the learnings from those activities form the basis for your future plans. Hell, sometimes just having one extremely experienced developer on a team can save you hundreds of thousands of dollars just by having them suggest the correct approach to tackle a problem.
One of these days I swear I'm going to quit my cruisy job and dive into this... one of these days.
