WAAS is more accurate than ILS at the outter marker
WAAS is easily more accurate than ILS at the outter marker.
Even at CAT I DH its still more accurate than ILS.
However after the DH, ILS becomes more accurate than WAAS.
But that's concerning today's single frequency WAAS receivers.
With dual frequency, triple constellation SBAS that should be ready to go by 2025, SBAS accuracy should be better than 2 meters in its whole core coverage area, not only in the final approach course !
2 meters accuracy will be enough for CAT II and CAT IIIa.
And for more efficient enroute / approach spacing, such ultra high accuracy positioning added to ADS-B should allow for 500ft vertical spacing at any flight level, since SBAS altitude actually gets more accurate at higher altitudes (more satellites in view), at FL 300, 30-45 satellites would be in view from a combined GPS+Galileo+GLONASS+SBAS GEO network.
WAAS makes is soo much easier to precisely and quickly join a virtual localizer and glideslope at any distance from the runway, as the SBAS receiver can tell exactly how far you are from the extended centerline, glidepath, and your PRECISE groundtrack course, no need to feel the wind. Something like 0.29nm left, 93ft low, 2.3 degrees intercept course (updated at least twice every second). Even with regular GPS, all that information is available, its just not as trustworthy as with SBAS.
The main issue is the politics behind SBAS / GBAS discourages enhancements to SBAS that make it better than GBAS ! Since billions of dollars have been invested in GBAS, there's a large vested interest in GBAS, also, SBAS is only viable if deployed on at least a continental basis, national barriers conspire against SBAS worldwide adoption. Finally the big plane / small plane separation mentality also discourages development of single/complete solutions that can be used on any size aircraft. The next generation of navigation gear should unify equipment design, so that a single, low cost solution can be used for any size aircraft, merging the SBAS sensor with the FMS, allowing for economies of scale to bring most FMS features into all IFR aircraft.
SBAS was designed to work with just 5 GPS satellites in view, while with todays full GPS constellation, 8-12 satellites is common, add just a full Galileo constellation, and 16 satellites minimum will be the norm. That's akin to a 500% safety margin improvement.
At the same time all SBAS limits are based on theorethical worst case mathematical calculations instead of real world results about 4x better ! If the FAA were to use just worst case real world results, CAT II would be available today with current SBAS (even on a little spam cam). But they have no incentive to make that happen. Too many economical interests would be hurt if that were to become reality.