I admit that I've never been much of a fan of SMS surveys. Too many limitations to do anything meaningful. Once Nathan Eagle started showing up at conferences and talking about his experiences with SMS in developing countries, especially in Africa, I could see a niche. That is, if you can call accessing two billion people a niche.
Last week someone sent me this little news item about a study by two academics, Fred Conrad and Michael Schober (I've collaborated with these guys on some unrelated work), that suggests one can collect some pretty good quality data via text. While you can't collect a lot they found that respondents were generally honest (no sign of social desirability bias on sensitive questions), showed little evidence of satisficing and exhibited few signs of heaping, that is, giving round numbers (typically numbers divisible by 5) when asked to report specific counts such as the number of movies they saw in the last month.
Given the kind of research that dominates MR I still believe that mobile ultimately depends on the continued growth in smartphone use. But whether via smartphone or good old-fashioned feature phone, feedback by text message may not be dead yet.