For fixed, point-to-point installations there is a special rule in effect: you can get more EIRP by having higher gain antennas as long as you back off the power; e.g. you can drop down to 22dBm (159mW) but have a 30dBi antenna and get 52dBm (159W) of EIRP.
A 30dBi antenna at 2.4GHz is, by the way, something like a 1.5 meter diameter parabolic grid.
E.g. I followed the link to '802.11 wireless LANs', thence to to '2.4_GHz_radio_use' where I was surprised to learn that US 'Amateur Radio operators can transmit [video&audio] ... in the 2.4 GHz band—and all ISM frequencies above 902 MHz—with maximum power of 1500 watts.'
But that's 5GHz or 60Ghz.
You could emit 500w in the 2.4GHz ISM band legally as a ham radio operator, as there is an overlapping ham allocation, so long as you complied with the ham requirements (which limits you somewhat on the 'whatever modulation' front).
But why would you? You're limited to line of sight communications -- far less than 500w will get you to the horizon. :P The only application I can think of for 500w at 2.4GHz would be moon-bounce. :)
Outside of the ham radio overlap, any intentional transmission in that band would have to be via a type-certified device. ... So it's more a question of what devices the fcc will certify than what you're able to do.
Combination hunting rifle and microwave oven: cook your ducks on the wing!
I suspect it would take unfeasible amounts of power to pull off in reality. Pesky humidity.