I agree.
\ufffdThe airline experience has shown us that it is not just technology that provides safety but the maturity that comes from a high-level of flight activity,\ufffd Rutan said.
Ideally the control algorithms would be based on the knowledge of the plane designers, the world's cartographers, the best knowledge of the atmosphere, and decades of man-years of pilot experience. That stuff doesn't come together fully formed like [link|http://www.ibiblio.org/wm/paint/auth/botticelli/venus/|Botticelli's Venus]. It's an iterative process that takes time.
There are plusses-and-minuses in being able to turn off FBW, or letting the pilot have too much control too. [link|http://www.aviationexplorer.com/airbus_accidents.htm|Here's] a list of Airbus accidents:
22 March 1994; Russian International Airways A310; near Novokuznetsk, Russia: Lost control and crashed after the captain had allowed at least one child to manipulate the flight controls. All 12 crew and 63 passengers were killed.
:-/
Many of us probably remember the Controlled Flight Into Terrain (CFIT) in the Everglades of Eastern [link|http://en.wikipedia.org/wiki/Eastern_Airlines_Flight_401|Flight 401]. Though in that case it sounds like there was, again, a control ambiguity issue, I don't think that argues against autopilots, for example.
Cheers,
Scott.