1. 11.8 years is a halflife, not a "all the methane is gone after 11.8 years" lifetime.
2. Methane doesn't just warm the atmosphere up a little bit and then disappear with no side effects. In addition to carbon dioxide, methane decomposition creates ozone and water vapor, which are both greenhouse gases. The additional heating effects of these decomposition byproducts are also included in the global warming potential calculations.
3. We care about cumulative effects over time. GWP is "how much additional heat will the atmosphere absorb because of this gas over X amount of time", scaled relative to carbon dioxide (so CO2 always has a GWP of 1). Methane's GWP-20 is about 80, which means that if I release one ton of methane today, over the next 20 years it will absorb about as much heat as if I had released 80 tons of CO2 instead. The longer the time frame the less bad methane looks, because it mostly decomposes, but even over a 500 year time frame releasing 1 ton of methane absorbs as much additional heat as if you had released 10 tons of CO2 instead. GTP is similar to GWP except it's about how much global average temperatures will rise instead of how much heat is absorbed.
4. If you can create methane out of atmospheric CO2 for free, you can subtract 2.75 from each of the GWP numbers for methane (since you remove 2.75 tons of CO2 to create one ton of methane). This is essentially what the table is showing on the CH4-non fossil line (notice each of the GWPs on this line is 2.8 less than on the CH4-fossil line).
-----
Imagine I had a magical gas called timelockium. It is not a greenhouse gas (no radiative forcing), but after exactly 10 years it decomposes to an equal mass of CO2 with no other byproducts.
The GWP-10 for this gas would be zero: over the first ten years, releasing a ton of timelockium is equivalent (in terms of heat absorbed by the atmosphere) to releasing zero tons of CO2.
The GWP-20 for this gas would be 0.5: over the first twenty years, releasing a ton of timelockium is equivalent to releasing 0.5 tons of CO2. This is because it does nothing for the first ten years, and then for the next ten years it is just CO2 [1].
For longer time frames, the GWP of timelockium would approach 1. Over 500 years, emitting a ton of timelockium would be nearly equivalent (0.98) to emitting a ton of CO2.
---
Now I have another magical gas, decayium. It is equivalent to CO2 for 10 years and then magically disappears. Again it has no other side effects or byproducts.
The GWP-10 of decayium would be 1--over the first 10 years it's identical to CO2. Over the next ten years it contributes nothing to warming, so the GWP-20 would be 0.5. For longer time frames the GWP of decayium would approach 0. the GWP-500 would be 0.02.
---
Superdecayium is like decayium except much worse. It's equivalent to 100x as much CO2 for the first ten years and then magically disappears with no side effects or byproducts. The GWP-10 is 100. The GWP-20 is 50. The GWP-500 is 2.
This last scenario is more analogous to methane, except methane chemistry is much more complicated, with gradual decay and byproducts that are also greenhouse gases. Like superdecayium, methane's GWP decreases over longer time intervals, but even over 500 years it is still worse than an equivalent mass of CO2.
---
[1] For the sake of simplicity I'm ignoring CO2 dynamics here, assuming it's just static in the atmosphere.