Andrew Gideon wrote
In his book "WeatherFlying," Captain Bob Buck offers this advice:

First, fly from good weather to good weather; ceilings should be 1000 feet
or higher, tops 7000 feet or lower. Takeoff and landing, of course, are in
VFR conditions.

Second, bad weather to good weather...take off into an overcast, climb to
on top, land in VFR conditions.

Third, good to bad...take off VFR, shoot an approach in deteriorating
weather. If you can't get in, you can always turn around and go back to
good weather.

Fourth, bad enroute. Take off in decent VFR, fly in nasty conditions
(clouds, no ice), land in VFR conditions.

His fifth step deals with flying in thunderstorm weather, and I draw the
line at that.

At a minimum, I'd put #4 before #2. #2 might require a quick approach to
return in case of a problem.

I strongly disagree. Option #2 puts you in weather for only a few
minutes, and at the very beginning of the trip when the weather
information you have is at its freshest and your weather picture at
its best. Relatively little is likely to go wrong, and if it does you
have an escape path to VFR that you have already planned.

Yes, option #2 makes less of a provision for mechanical failure - but
mechanical failure is far less likely than unforecast weather
deterioration.

Option #4 puts you in weather for prolonged periods. There is every
opportunity for the weather to deteriorate without you catching it.

I just flew an example of #4 this past weekend, and it was easy. I took off
in VFR, climbed into clouds, spent much of the trip in clouds, and dropped
below the ceiling to execute a visual approach at the destination.

Plenty of airports along the way were VMC, so I'd plenty of options.

I've seen options disappear quickly and over a wide area. Sure,
option #4 is TECHNICALLY less demanding - it requires less skill in
aircraft control and following procedures. But that's precisely the
area where the instrument rating curriculum is strongest. Option #4
also puts the most demands on your ability to monitor and predict the
weather - exactly the area where the instrument rating curriculum is
weakest.

Michael