LH2's comment more fully was - Barometric pressure corrected for the local aircraft effects through the ADC (i.e. the 'consistent' pressure level) is used for RVSM because it is generated internally and has no requirement for reference outside the aircraft. In addition, the outside air pressure is something that can be measured quite accurately with multiply redundant pressure sensors and therefore has a high degree of resilience.
Actually, that's not correct at all. Barometric pressure, corrected through the ADC is used for RVSM, and relies on reference outside the aircraft. It's barometric pressure, after all.
So far as accuracy, it's a subjective term. Barometric altimetery is accurate enough for RVSM, to be sure, but certainly not accurate in revealing true altitude. One's actual height is largely irrelevant when flying a barometric altitude or flight level, save for obstacles which couldn't give a whit about barometric altimetery.
As for resiliency, one may look to hysteresis for repeatability, but even in advanced systems, I see errors that are hardly resilient. I recently ferried several airplanes that had been sitting in the desert for a year or so, only to find that altimeters disagreed by several hundred feet at altitude. Many of the older Lears that I used to fly were ADC corrected on the captain side, but pure baro on the copilot side, and were nearly always at least 500' different in their displays. That the display might show this or that on any given day was always a matter of guesswork, as well as temperature and airspeed, for without correction and input, the baro altimeter was anything but resilient.
Improperly set, the baro altimeter is anything but forgiving.
Why then, is GPS not the golden standard for setting flight levels and maintaining them in the compacted, high speed world of the upper flight levels? If it's so much better at providing an altitude, as some suggested, why is it not used?
A barometric altimeter may freeze; static ports may seal off, and many utilize vibrators to ensure minimal hysterical errors (a Freudian discourse in it's own rite, I'm sure), yet they continue to work in the absence of generators or heat with as little effort as opening an alternate static port. The same can't be said of a GPS, which is always dependent on mother-battery, or father-generator for life.
Barometric altimeters have been used for many years and will continue to be used for many to come, for good reason, in the same plain, old fashioned way they've always been used. Why? They work.