Have had this discussion before, though not recently.
The most basic, most often correct answer when flying a piston twin is positive rate, gear up. The correct answer when flying a twin turboprop or jet, for that matter, is the same. Positive rate, gear up. But if someone were to say, "Well, in MY twin the right thing to do is to pitch for best rate before retracting the gear," they may very well have some backup in the form of manufacturer guidance.
Checklists are partially to blame for this confusion. Some simply state "positive rate / gear up." Others say to attain best rate of climb speed, then retract the gear. Others say to retract the gear when no usable runways remains. As is often the case in aviation, the lawyers often have more to say about the checklists than pilots do.
But looking at this from a practical perspective...
In a light twin, you're looking to reach Vyse as soon as possible. Especially in underpowered light twins (Seminole, Apache, etc.) leaving the gear down increases the amount of time required to achieve Vyse, so why would you leave the gear down, particularly when there's no remaining usable runway?
The corollary here is that "positive rate, gear up" can still benefit, particularly in a training environment from the addendum of "... when no usable runway remains." However, in my experience a disconcertingly large number of pilots misjudge localizer antennas, fences, trees, cars and houses for usable runway... YMMV.
But waiting for Vyse with the gear down, particularly at high DAs, heavy weights, flying sluggish normally aspirated light twins, is pretty dumb once you've got no chance of landing on any remaining runway, and flying off the airport property with the "Dunlops deployed" is just egregiously stupid. There's no conceivable benefit to waiting for Vyse in that case, and all one could possibly be doing is delaying that important safety milestone airspeed with extra, useless drag.