I have often stated that some structures have zero seismic capacity. That is not the same as zero ground motion, since any structure must stand up to the live loads such as wind and people dancing.
As a recap, I have dedicated my whole life to eliminating acceleration in reference to seismic. This was a standard adopted many years ago because accelerations are easily measured, and can be used as a lateral body load, as a percentage of the g force. In fact, as a design coefficient it is adequate. A building designed to 10% of lateral load (a slight tilt), is quite robust. The building codes all use this coefficient method.
But it has nothing to do with actual seismic measurements. The impact of acceleration on buildings goes down to nothing as the frequency increases. To continue to use peak acceleration (PGA) in relation to actual measurements and seismic hazard is pig-headedness, that I fear, will outlive me.
Those with brains in their heads have gone to peak ground velocity (PGV) which scales perfectly with frequency as it is a measure of induced strain in both the ground and the building. You can easily use a velocity time history in a more authentic analysis using finite differences.
Enough about that. What is the seismic capacity of the worst building in the world? I would assume that it is composed of stacked bricks, using brick arches and no mortar. This is common throughout the world for older buildings. These are the types of buildings that kill massive numbers of people for their 1 in 500 years earthquake.
I will state that the seismic capacity to death is 5 cm/s, for these things. We must expect soft soil to amplify PGV (not PGA!) by a factor of 10. That's somewhere about 3 MM intensity units. That means the firm-ground motion is 5 mm/s. You get quite disturbed at 0.5 to 1 mm/s. That's what hits these poor Oklahoma people all the time. The blasting standard is to keep things below 5 cm/s, which is the main reason I use it.
This latest Afghanistan earthquake was over 200 km deep. The ground motion on firm ground (if there is any over there) could not have been over a couple of mm/s. Some people got injured and killed, which is expected for extreme amplification (over a factor of 10).
So, my 'zero capacity' is 5 cm/s. Moderately designed structures should take 50 cm/s, and I'd be happy at 1 m/s for soil. If you are solid rock, you can never get more than 1 cm/s, unless you are right in the near-field of a shallow thrust, on the hanging wall. Bridges probably can't stand if they are less than 20 cm/s, and you can build all you want at 5 cm/s, as long as you don't expect to live more than 500 years. :)
But as a rich society, I don't think we can afford chances at 1 in 500 for total destruction. We have seen city after city destroyed at these odds. For an individual city like Toronto, you can laugh and build your house of sticks, but the Big Bad Wolf does come around. Toronto would be destroyed with a thrust at 1 cm/s because of all the soft soils. As well as a fine nuclear disaster at Pickering. :) Would we care about a little extra dose? Really, what was marginal death count for radiation in Japan, as a percentage of total deaths?