serious error (re. radiation shielding) in 1975 Summer Study?

I've been working this week on modeling the radiation environment for
various orbits, and the effects of shielding. You can see my notes here
[1]. My results differ substantially from the claim made in the 1975
Summer Study (and subsequently oft-quoted elsewhere).

The claim is made (in Chapter 4 [2]) that 4.5 t/m^2 of passive shield is
sufficient to bring GCR (galactic cosmic radiation) dosage down to 0.5
rem/yr. The only justification given for this is a comparison with
Earth's atmosphere, which is about 10 t/m^2.

But right off the bat we have a discrepancy here: Earth's atmosphere is
(obviously) within LEO, and by that point, most of the GCR has been
deflected by the geomagnetic field. I found several references in rough
agreement that a craft in LEO receives only 1 mrad/day of GCR, which
works out to about 0.36 rem/year (using the approximately-true
assumption that 1 rem == 1 rad). In comparison, free space (including
L4 and L5) receives not 0.36, but 60 to 80 rem/year of GCR.

So it seems obvious that, considering passive shielding alone, the
Earth's atmosphere would be woefully insufficient outside of LEO. Much
less *half* that much shielding.

And this agrees with my results... starting with a GCR level of 65
rem/year, and an attenuation function roughly fit from Rapp 2006 [3], 5
t/m^2 of shielding gives me a GCR dosage of 19.1 rem/yr. This is much,
much too high according to the 0.5 rem/year goal set in the Summer Study.

It gets worse. As Rapp points out, you get diminishing returns from
more shielding, as most of the low-energy particles have already been
blocked and the high-energy ones continue to penetrate. When I increase
the shielding mass to 20 t/m^2, I still estimate a dosage of 1.8 rem/yr.
To get under 0.5 rem/yr, I need about 50 t/m^2 of shielding ten
times what the Summer Study claimed.

Now, I hope I've screwed up somewhere. It's certainly possible that my
rough curve fit to Rapp's results is incorrect, or even that his results
or figure contain an error. I'll look for other sources, and also
double-check all my math today to make sure I haven't slipped a digit.

But the basic "sanity test" check on the Summer Study results bothers
me... how could half as much shielding as provided by the atmosphere
suffice in an environment that receives 200 times as much radiation to
begin with?

Of course this was a 10-week study done decades ago, and I'm sure there
are errors in there, but unless I'm overlooking something, this one is a
real doozy. Radiation shielding dominates the mass of any habitat, and
if we've all been off by a factor of ten, this seems like big news.

What do you think?

Thanks,
- Joe

[1] https://www.assembla.com/spaces/high-frontier/wiki/Radiation
[2] http://settlement.arc.nasa.gov/75SummerStudy/Chapt4.html
[3]
http://www.marsjournal.org/contents/2006/0004/files/rapp_mars_2006_0004.pdf