I did a quick google.
When radio waves propagate through the ionosphere they suffer an extra time delay. This time delay is a function of the total electron content (TEC) of the ionosphere. Since the ionosphere acts as a dispersive medium to GPS signals, dual-frequency (L1 at 1575.42 MHz, L2 at 1227.60 MHz) GPS receivers can eliminate (albeit to the first order) ionospheric delay through a linear combination of L1 and L2 observations
Single- versus Dual-Frequency Precise Point Positioning
The improvement is only to the first order (i.e. not to the level of differential GPS where you rig up a GPS station on the ground) but that is way better than what is achieved in practice with WAAS/EGNOS. OTOH it appears that the computation needs time and there is where the tradeoff is if the receiver is continuously moving.
Looking back at the first post above, it sounds like the Galileo approach is to use E1/L1 + E5/L5 whereas the US Navstar approach is to use L1 + L2.
