The bubbler would need to be able to withstand high temperatures, but the water -when presented to the manifold pressures- would vapourize that much more easily.
However, the bubbler would be acting as a sort of catch can for everything burned and unburned in the exhaust stream, as the source for EGR gasses is pre- catalytic converter, so it would get dirty.
I was thinking about this too, so after my last reply, I revisited my water bubbler set-up. The problem I had before was with modulating the amount of vapour, so I added a venturi (needle) valve to the port on the intake George Wiseman suggested I use. I don't have a photo but here's a diagram:
at low throttle angles when manifold vacuum is high, water vapour gets ingested by the engine mostly from the bubbler.
at higher throttle angles, when vacuum is closer to atmospheric and the bubbler isn't as productive, blowby vapours from the PCV partially make up for the drop in bubbler production, as the system was meant to. The needle valve modulates the amount of vacuum applied to the bubbler, and I've set it to just start a "rolling boil" at hot idle, so water vapour starts collecting on the inside walls of the vessel.
I've driven ~125 miles so far on just over 2.5 gallons (that's a solid 50MPG) according to my dashboard readout...and it's fairly accurate when it comes to comparing how much it pumps to how much I have to put in (within a few percent of what the station pumps tell me, which I chalk up to instrumentation error on the part of both)