This is a fantastic way to look at Tweetbot versus the Twitter app, Kaishin L.Reda explains:
To see if Tweetbot lives up to the hype, I tried to pit it against the official client using a GOMS-inspired, oversimplified human information processing model (HIP) on a set of frequent as well as less frequent tasks we perform on Twitter.
Which basically means that a time value was assigned to every action (less typing time) and then each app was put through its paces. So for instance how quickly can one reply to a tweet in each app. This is a very neat method and the results are basically a wash.
The more interesting conclusion that was drawn is that it would seem Tweetbot is better when used as a Tweet reader, rather than that of a conversation tool. Meaning: you read tweets more frequently then you actually send tweets.