Not after takeoff??
Instrument rating renewals ALWAYS included a simulated engine failure after takeoff. The stipulation was "not below 50 ft agl, and not below VMC plus ten knots with gear up"
But it was usually below max takeoff weight. Usually full fuel and two POB.
I did a bit of research with the Partenavia when we first got them. I found that they would climb a little on one, if you fly them straight, and at the right speed. But in a rate one turn they would lose about one foot of altitude for every degree of turn. (you do the sums)
So it is important that you fly them right and have an escape route that does not require immediate climb or turns. Then when it fails you climb straight ahead for as long as it takes to get a height that will tolerate a turn. A "minimum assymetric circuit height" You WILL lose some height in a turn in most piston twins.
And don't think that your whizzbang twin will perform so well that this reasoning does not apply.
There was a Kingair that went assymetric at Sydney and could not maintain height for a runway that was 20ft amsl.
You DO have a little climb performance, but it is fragile and you can destroy it easily.
The same applies to controlability.
Assymetrics sort out the men from the boys.
I remember reading american research that showed that you are less likely to be killed due to an engine failure in a twin than in a single. But you are more likely to be killed in a twin than a single.