Wait until dual frequence GNSS is widespread
The big limitation about SBAS today (WAAS, MSAS, EGNOS) is the iono grid.
You need lots of reference stations to calculate a precise enough iono grid.
That's why there are no LPV approaches in Hawaii, even if the FAA were to add one station in every major island plus one in Guam, the Marshall Islands, still the iono grid calculation wouldn't be good enough.
But all of this ends with dual frequency GNSS.
Dual frequency receivers calculate iono corrections locally, more precisely than a centralized grid broadcast. For aviation that means L1 and L5 bands (the dual frequency pair).
Once that is true, it would take just 4 MSAS stations in Australia (perhaps one in New Zealand) to have full LPV coverage for the whole region.
All you need is one station in the extreme SW, NW, SE and NE corners of the region and you're done. While to do single frequency MSAS you would need a dozen stations for full coverage. The sole reason you need more stations is so satellites coming in view over antartica (that aren't in view of stations in Japanese soil) can be tracked properly, and have their ephemeris and clocks corrected. For that you need to do the reverse of what the GPS receiver does (get the signal on 4 stations with good geometry and compare where the satellite says it is vs where it actually is).
Finally, having MSAS support for Galileo will be a big part of this, cause they come with the L5 signal standard, while GPS is very slowly launching new satellites with L5 support. It's likely dual frequency SBAS will become doable mostly due to Galileo, as full dual frequency GPS coverage (at least 24 satellites) is expected to take at least until mid 2020s (if you use actual launch rates instead of the US Air Force way too optimistic schedules).
But with half GPS L5 coverage plus full Galileo coverage it would be arguably possible to even go CAT II approaches with SBAS.
Using full Galileo + incomplete GPS L5 support, enough coverage could be in place by late 2017 (assuming about 24 months delays over the current Galileo schedule).