There is no regulatory regime, nor IMHO will there "ever" be, in which synthetic vision will be regarded as VMC. The terrain data comes from the US space shuttle (SRTM) radar imagery, done some years ago, which (in its publicly released version) is accurate to about +/-100ft (though it has been degraded outside the USA - I recall reading the SRTM website mentioning +/-300ft) and I guess that for regions where there is "enough business" (the USA in this case) they supplement it with the obstacle data.
I've got a Garmin 496 yoke-mounted and this has the SRTM database for Europe; this is my poor-man's EGPWS but is actually really effective. It has European obstacle data too, and so far it has been found to work in all tests against terrain. I get the TERRAIN AHEAD PULL UP warnings in the headsets, just like the "certified" version costing thousands.
But when it comes to SVT, having seen some stupidity (like what LH2 refers to; the utterly trash terrain data in Flitestar) in the way some of the data is managed, I would not trust my life with it for casual flight in IMC. The GPWS technology is wonderful as a last-resort lifesaver in cases of a total loss of situational awareness, which is why big jets all have to have it (and it has reduced the CFITs to more or less zero) but that's a different thing to doing a whole flight with it, below the MSA, relying on the database accuracy all the way along. It's a difference between doing an flight from Southend to Trondheim carrying a life raft, and doing the same flight with the spark plugs done up only finger-tight.
Especially in Europe where the GA scene is much poorer and consequently the database will get so little real testing. It's like flight planning programs - you can have glaring errors in some southern European countries and they stay there for years, because there is almost no GA in these places. So you take your new Cirrus with SVT for a trip down to say Greece and you will probably be the first person ever to be exercising the system down there. Think about that! For flight in IMC below the MSA? Not me - I've been in hardware/software for many years
I often write negatively on new developments in GA avionics but that's because I've seen some of the utter crap that is out there, developed by people who have little experience of developing robust software. Certification (TSO etc) means nothing in GA because the manufacturer simply ticks boxes on the application form and sends it off to FAA or EASA. In the end, the product is only as good as the testing it gets in the market, and only as good as the market feedback gets taken notice of (probably more by Garmin than by e.g. Honeywell), and in "poor Europe" there will be so little feedback that
you will be the guinea pig for everything new. The Q is: do you feel lucky?
There is also an ongoing cost. The databases are not cheap. Look at the cost of updating the terrain database in the G496 and multiply it a few times. Add in "nice" stuff like Jeppview approach charts and georeferenced airport diagrams (which is infinitely more useful) and you could easily be looking at £2000 a year in updates, just for Europe.
Nice product though! Long time coming, since MSFS had this for how long? 10 years? You can buy X-Plane with the world's SRTM scenery on half a dozen DVDs for about £50.