I think that simulated engine failures in situations described in the JWW accident report and NEJ accident report (Tamair Metro) should be authority enough to show that it is unsafe to perform that exercises at night below circuit height - especially in the case of a light training twin with marginal single engine climb performance.
In both of those accidents the aircraft had take-off power being delivered on one engine, this has the obvious effect of a large yaw force that requires a quick response to correct.
Of course, a simulated engine failure during, say, a turn from an outbound leg to an inbound leg on an NDB approach or at the outer marker of an ILS is completely different, due to the active engine only delivering approach power. The yaw force is much less and therefore there is a greater margin of safety (although the effects of an engine failure during this phase of flight can be insidious).
I would be interested to read the accident report regarding CTT, what factors were contributory.
For those interested, here is the link to the ATSB report of the NEJ accident:
http://www.atsb.gov.au/aviation/pdf/vh-nej.pdf