Without data of any kind, this is merely a hypothesis. I do not buy it; it assumes all mental processing is serial rather than parallel, and takes no account of the fact that as someone types L-R in a text field, the momentum of rightward-moving attention is quite different from scanning static UI elements. Also, it's not hard to see a text field and search button as a compound single UI element, obviating the need to visually navigate it each time.
I might feel differently if I used my left hand to operate the mouse, but then again I might be used to it by now.