I really don't want to be Debbie Downer, but WOW, that is a really small sample size over a very short period of time to extrapolate anything conclusive.
ComScore says there were 63.2 million smartphones in the market in 4th qtr 2010. I'll do the math for you, (just had to relearn "e-5", thanks Google) and there you have it, .00001% sample size. A 3 month survey when the avg time to sale is at least 6 months seems insignificant to even get good data from that sample. Then there is the issue that Uncle Joe will bring up if I don't, "What did they ask?" I always want to see the questions in a survey like this, but especially when the group that publishes the study has a lot to gain from the results. Here is a quote from the article:
"Greystripe said spending from the category on its network more than doubled in the last year. The company doesn't break out spending by industry, but said automotive is one of the largest categories on its network."
I'm not saying to stay away from mobile marketing, but I don't think this study should be the ONLY reason you decide to jump in headfirst either. I'd bet that you can get better data from polling the people that actually step on your lot with your own questionnaire.