Sorry, haven't read the thread so not sure if this has already been raised.
Car and Driver did a test and found that texting did impact reaction times, sometimes by a CRAZY amount.
http://www.caranddriver.com/features...angerous-is-it
The sample size is way too small to say anything other than it is suggestive. They should do this with a larger sample size.
But if that is valid I like the comparison with driving under the influence and I think it's somewhat valid. Not sure if more testing has been done, but I assume that there are some stats available on after-the-fact accident causes that quantify texting?