I understand there are two problems with magma and heat: (1) how to avoid the ridiculous outcome of having the magma pipe melt through everything, since the obsidian it's encased in should melt. (2) how to avoid having your dwarves catch on fire automatically every time they try to use a magma forge.
About (1), this actually shouldn't be a problem if heat flow and specific heats are handled properly. (I know specific heat at least is in the raws, but I don't know if the behavior is correct.) Near surface magma is almost always at temperatures only slightly above the temperature at which it will cool into something solid like obsidian, while most everything on the surface is significantly cooler; as a result, you can get obsidian quenching around almost anything at the surface, even highly flamable things like trees:
The lava/magma that next encounters what the first wave quenched against needs to melt or break that obsidian shell before it can progress any farther, resulting in lava/magma that only spreads out until the heat flow at its edges becomes too low to stop the formation of obsidian walls that the magma doesn't possess enough pressure or temperature to overcome.
(2) is a bit more difficult; my hunch is that Toady's earlier experiments involved constant (or heavily buffered) magma temperatures, or that air and rock was taking heat from the magma too quickly for the simulation to dissapate it in a realistic fashion. Or maybe dwarves and other creatures need some intermediate level of being burnt between "feeling good" and "OH GOD MY BEARD IS ON FIRE". (I wouldn't be suprised if Toady planned burn scars with this release anyways; it would be nice for magma forge workers to get burns and singed beards just through their normal working conditions. Other materials should probably burn and singe too so that their clothes don't suddenly go up in flames.)