I'm combining stuff from both replies, bear with me.

Add 20 hours of aircraft time (minimum, in my estimation, because of
all the rehash, and startup overhead) and you've added $2000 right
out
of the shoot.

First, I don't agree with your cost asessment. Around here, an
instrument trainer rents for $60-$80/hr. Between airline tickets and
10 days of hotels and meals, you're looking at $1500, easy. So even at
20 hours, the costs there are a wash and my original asessment holds.
In areas where the rentals cost more, hotels and meals do too.

Second, I don't agree that 20 hours is a minimum - more like a maximum.
I completed my rating (in the non-accelerated mode, stretched out over
half a year) in 43 hours, and my FIRST student (I would like to think
I've gotten better since then) that I took from zero was done in under
55 - despite major equipment problems, the inefficiencies of
structuring the training to get what actual we could, and having the
process stretch out over more than a year. Had I been willing to
ignore opportunities to get actual, and had we not had several sessions
where the glideslope had problems (how would THAT have affected an
accelerated course?) we would have been done in well under 50 hours.
Also, his direct operating costs were about $25/hr (Pacers are cheap to
fly).

Third, I would go so far as to suggest that most pilots who need 20+
hours more to complete the rating flying once or twice a week rather
than on an accelerated basis probably won't be safe once they get the
rating. If they forget so much week to week, how much will they forget
when they go weeks between approaches?

You are not considering difference in effectiveness of the training
device. The Frasca blows any aircraft away, in my opinion, (and I've
done it both ways) as an efffective and efficient learning tool.

That's true if the training we're focusing on is scan and procedures.
Of course everyone is different, but I found that even in the airplane,
I was proficient at scan and procedures prior to the 20-hour mark. Of
course scan and procedures are essential for safe and capable IFR
flying, but they are far from sufficient. The real issues are ATC and
weather, and those can't be learned on the simulator at all.

Non-accelerated, you have a rating in 8-12 months.
Accelerated, you have the rating in 10 days, and spend those same 8-12

months flying in the system and gaining experience.
Who's thebetter instrument pilot at the end of those 8-12 months,
would you suppose?

That depends - did the student who did the accelerated course learn
enough to be capable of flying weather and learning further on his own?
I'm seeing an awful lot of students who seem to need an instructor
when the weather goes bad. To me that indicates a problem. Because
weather is what it is in Houston, I am generally only able to get my
student about 5 hours of actual in the course of training (and believe
me we make it a point ot get it if it is available, even if it's not
the most efficient way to get to the checkride) but they're all able to
go out and fly weather on their own.

If the accelerate training employs good instructors, I don't see why
those students should be any different - and thus you are right, of
course they will be the better instrument pilots. But if choosing the
accelerated program means settling for inferior instructors (and unless
you pay the premium for an outfit like PIC, it certainly will) then I
don't agree. The student who got inferiour training will not have been
progressing in those 8-12 months unless he was carrying an instructor
around in weather - in which case, what was the point of having the
rating?

Like I said - I'm not saying a program like PIC isn't worthwhile,
merely that you will pay a premium for it. And if you replace their
multi-thousand-hour instructors with standard FBO timebuilders, then I
would say it's not worthwhile at any price.

Michael