That said, P.S. Jone’s suggestion that literacy and usage intention are what ultimately define utility is spot on. With my Android, I regularly speak 3 to 5 page documents directly into Google Docs. It’s utterly annoying to tap out blog posts on a handheld portal; but voice makes the task trivially easy. I use voice to text with the Rooftop app to directly collaborate with our team’s Highrise database, and the Basecamp project management website; both exactly the same way that I’d do from laptop or desktop. Granted, bandwidth and screen real estate constraints do slow me down maybe 30% to 50% in some instances; but the comparison to zero productivity in the pre-smartphone era makes productivity gains glaringly obvious, IMHO.

The Econ 101 question is always, “compared to what.” Compared to dumb phones, a motivated and literate smartphone user is already a hyper productive cyborg such as the world has never before seen. Talk to @caseorganic about the how and why of today’s cyborg populations. ;-) In the palms of our hands, we carry extended cognition portals that link our minds in realtime and illuminate incredibly rich information, location, and contextual awareness layers upon the physical world. Without that portal, I’m a qualitatively different human being. Non-enhanced. Pre-cyborg. It might not be hyperbole to even say: crippled, handicapped.

Apps like My6Sense learn from my behavior and can discern personal salience amongst a proliferating diverse of activity streams, saving me hours of time that used to be spent scanning across reverse-chronologically ordered Google Reader filtered feeds (qualification: “sort by magic” in GR has improved that process, as well). Tweetdeck makes #hashtag topical awareness a realtime fingertip experience.

In terms of pure text output, for slowish typists, voice to text on a 4″ handheld portal can be materially more productive than hunt-n-pecking on a $195 ergonomic keyboard with screaming fast 3GHz desktop computer and 42″ LCD monitor. Because I’m a moderately skilled typist, maybe 70wpm or so, it’s perhaps 20-30wpm slower for me to use Voice to Text; but compared to zero, that’s a huge win.

Pardon the scatter-plot style response, if I had more time I’d back up, proof, edit, clarify. Instead, I’ll click “Notify Me” below and do my best to answer questions if they come in.

Perhaps we could think about doing some proper objective measurement of these types of every tasks to empirically measure the differential across some representative sample populations?

However, programs like Evernote have been awesome for helping me maintain thoughts/focus on the go if any good ideas pop up.