For instance, here are two glycemic curves after two different meals, "A"
and "B", with the same amount of carbohydrate.
Which food would have the higher GI in your opinion, A or B?

http://snipurl.com/glycemiccurveexample

Sorry, "Google Groups" is not an authoritative source of information. ;-)

Published studies are more useful in our case.

For instance,

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=pubmed&cmd=Retrieve&dopt=Abstra
ctPlus&list_uids=15523541&query_hl=7&itool=pubmed_docsum
Determination of the glycaemic index of selected foods (white bread and
cereal bars) in healthy persons.
[...]
METHODS: To determine the GI, measured portions of food containing 50 g of
carbohydrates were eaten by 11 healthy volunteers. B-glucose curves were
constructed from B-glucose values at time 0, 15, 30, 45, 60, 90, 120 min
after the meal. The GI was calculated by dividing the incremental area
under the curve (IAUC) for the tested food by that for the standard food
(IAUCS).
[...]

Full text
http://publib.upol.cz/~obd/fulltext/Biomed/2004/1/17.pdf

Of course, it can be said that  
"Carbohydrates that breakdown quickly during digestion have a high GI
because their B-glucose response is fast and high. Carbohydrates that
breakdown slowly have a low GI".

Still,
"The GI is a measure of the food power to
raise B-glucose concentration after a meal."

and GI is defined as
"relation of the incremental area under the B-glu-
cose response curve (IAUC) of a tested meal containing
50 g of digestible carbohydrates and the incremental area
under the B-glucose response curve of the standard food,
i.e. 50 g pure glucose (IAUCS)."

Here is how they determine the GI of a food in detail :

Determination of the glycaemic index

a) Getting basic data
To determine the GI, measured portions of tested food
containing 50 g of carbohydrates were eaten by each of the
11 healthy volunteers (Table 1) after an overnight fast; the
same approach was used after an afternoon fast. Finger-
prick blood samples were investigated at 15–30 minute
intervals over the next two hours after the meal (at times
0, 15, 30, 45, 60, 90, 120 min; the beginning of the food
intake was time 0).
Each volunteer measured his/her B-glucose concentra-
tions by means of a glucometer Optium. At the end of the
one-week test period the B-glucose values were transferred
from the memory of the glucometer into a PC for further
analysis.

b) Construction of B-glucose response curves.
The averages of the respective B-glucose concentra-
tions after the meal were used to draw a B-glucose re-
sponse curve for the two-hour period. The values at times
75 and 105 min were obtained by extrapolation.

c) Exclusion of disturbed tests.
For the purpose of statistical evaluation, all tests that
were not complete and all tests where the first (i.e. fast-
ing) B-glucose concentration was 7.0 mmol/l or higher
were excluded.

d) Calculations of individual GI values in every vol-
unteer.
The incremental area under the curve (IAUC) was
calculated for each meal in every volunteer separately (as
the sum of the surface of triangles and trapezoids between
the B-glucose curve and horizontal baseline going parallel
to x-axis from the beginning of B-glucose curve at time 0
to the point at time 120 min) to reflect the total rise in
B-glucose concentration after eating the tested food.
The IAUCS for the standard reference food (i.e. 50 g
of pure glucose) was obtained similarly to the mean from
the first three independent IAUCS1 , IAUCS2  IAUCS3
in the same volunteer.
In the IAUC/IAUCS calculations, all B-glucose val-
ues in the course of the test lower than the first value (at
time 0) were equalized to the respective first value.
In each volunteer, the GI (%) was calculated by di-
viding the IAUC for the tested food by the IAUCS for
the standard food and multiplying by 100. The following
formula was used:

[...]

IAUC – Incremental Area Under the blood glucose
response curve for the tested meal

IAUCS – Incremental Area Under the blood glucose
curve for the standard meal

e) Working out the average of GI’s for tested food in
each volunteer.
In each volunteer each food item was tested 5 times
so that 5 GI’s was obtained and (after the exclusion of
disturbed tests) the average was calculated.

f) Final calculation of the GI for each tested food.
The GI for each tested food was calculated as the mean
from the respective average GI’s of the 11 volunteers.

g) The variability of GI for each tested food was
assessed according to standard deviation of the mean;
histograms of GI values demontrated the frequency and
range of results.

==============================================