0.9DL is not used in EQ load combinations as per Table 4.1.3.2A. The load factor is taken as unity, despite the load counteracting any overturning. It doesn't mean you can't use 0.9D if you want to be conservative. This differs from some of the American codes where reduced dead load is used. My understanding is that NBCC's limit states actually use a statistically different earthquake than those codes (which may use MCE) and the Limit States calibration is different. Beware when using American software to check, as you may have to pro-rate the load externally.
In terms of vertical ground motion, no NBCC 2015 doesn't consider that at the moment. My understanding is that NBCC 2020 (2021?) or NBCC 2025 (?) will account for vertical ground motion effects.