Your link is very deceiving. they use facts, and terms invented by the IPCC et. al. that rely on the slope of a curve, that decreases as values increase.
Let's start with RE (radiative efficiency.) This is the slope created from the present level of a greenhouse gas by adding 1 ppb. It looks like this:
Please note, that the slope as plotted here is 0.0168 for CO2 and 0.4598 for CH4. I used 280 to 379 ppm for CO2 and 730 to 1774 ppb for CH4, and a slightly revised log curve so a zero value could be met. These values are close to the AR4 material, which even in it is inconsistent in values they use. The AR4 lists the RE of these gasses at 0.000014 and 0.00037 (page 212.) The factor of 1,000 difference is the slopes are because I plotted the graph in ppm rather than ppb. Anyway:
0.4598/0.0168 = 27.369
For the AR4:
0.00037/0.000014 = 26.43
So, this is the ratio of RE between the two gasses.
Now what GWP does is use this relationship, and assume an equal mass increase of 1 kg. See page 210 of the AR4. This means they are using 2.75 times more CH4 than CO2 to get the GWP number:
27.37 x 2.75 = 75.26
IPCC AR4:
26.43 x 2.75 = 72.68
Now that's the instantaneous number for the GWP calculation, then over time, the ratio changes dependent on the dissipation rates of the gasses. That's why we see the nominal 72 and 100 for GWP numbers. The link you provide is incorrect in that they say 100 times on a molecule by molecule basis, but it is mass. Not molecules counts that determine GWP.
You are citing agenda driven idiots!
Notice how fast the slope flattens as the gas levels increase.
Also consider this. The 1750 to 2005 forcing for CO2 and CH4
1.66 W/m^2 for CO2 from a 36.3% increase (278 to 379 ppm.)
0.48 W.m^2 for CH4 from a 146% increase (715 to 1774 ppb.)
Log curve fits:
CO2 W/m^2 = ln(379/278) x 5.35
CH4 W/m^2 = ln(1774/715) x 0.528
The ratio of 5.35 to 0.528 suggests that CO2 is 10 times stronger than CH4, bu CH4 is not a simply log fit since N2O and CH4 overlap too much in their spectra. The formula the AR4 uses for both gasses are closer to the graph I made than what a pure log formula yields.