Just an observation here. I find it interesting that those projects that are of immense complexity, such as huge multistory buildings, can't be done by hand. They would take too much time and hand calculations would introduce too many chances for error. The projects that need accuracy the most are the ones that depend on computers the most. And the future of structural engineering doesn't look like it's going to get any simpler (just look at the wind load requirements in ASCE 7-02!).
So, if it's good for goose, it's good for the gander? Why accept the computer stuff on the big projects without question but question the stuff on the little projects? Seems contradictory to me.
I personally do both - hand and computer (including purchased software and spreadsheets that I have developed in Excel). I would use more computer programs if I could afford them!
I remember when we were first allowed to use calculators in high school. My father would not let me use my calculator at home without doing the problems by hand first. Today, calculators (and I'm talking fancy graphing things) are required by high schools, even down to a particular model and manufacturer.
This debate almost sounds like a cultural thing. The old guys (like me) grew up with slide rules and calculators that might do square root. Kids today are growing up with technology that boggles the mind. They have implicit trust in this stuff to the point that questioning it's usefulness is like questioning the value of a car (duh!).
We need to accept that fact that technology is here to stay. As a community, we should be developing training methodologies for new engineers that utilize these tools effectively instead of forcing hand calcs as a check against the "machine". At the same time, we need to lobby the software companies to develop programs that we can use instead of what makes them money.