I am no fan of the global warming alarmists and the series of modifications they have made to the temperature record of the last 100 years or so. Usually you change the theory to match the data, not the other way around.
I watched an old Nova program tonight about the Siberian Traps. The people studying the Traps claimed that the outgassing of CO2 from the massive amounts of lava caused the temperature to go up 10 degrees (Fahrenheit, I think, but I don't remember for sure), and the lava caught the large coal beds in this region to catch fire and the consequent release of methane (a stronger heat trapping gas than CO2) caused temperature to go up another ten degrees. The resulting warmer oceans killed a lot of the sea life according to their theory.
I imagine the time scale of your temperature and CO2 plot is not fine enough to show a relatively short spike in CO2 if it did happen the way they theorize. They believe the CO2 concentration reached three times the current level. They used fossil ginkgo plants from the period of the Traps' formation to estimate CO2 concentration. Ginkgo leave apparently have special pores to absorb CO2, and the researchers think if there is a lot of CO2 present the plants didn't need as many special pores. So they counted special pores on the fossil leaves.
I don't know what to make of their theory. I'm suspicious of their explanations, but the immense size of the lava flows may well be related to the mass extinctions.
Your plot seems to show an inverse relationship between CO2 concentration and temperature in places or perhaps no consistent relation between the two.
I know that rising temperatures in the ocean would cause some methane clathrates and CO2 clathrates to melt in previously cool ocean sediments, but that may be a pretty slow process. Too slow perhaps to cause a runaway outgassing or explosive release of methane and CO2.