It is my current understanding/belief that standardized testing would show *negligible* differences in HP (Power) and RPM between two engines with differences in mass being the only variable.
That being said, I choose to interpret Victor's: "more top end... in some situations," as a real world practical consideration. MORE, not HIGHER.
With the same Power, you would need less time to get things up to speed. and because you usually brake to stop instead of coast, that means MORE time using power for propulsion instead of mass investment.
and the bike's weight will be marginally lighter, which again increases acceleration. which is more TIME going fast. which can be interpreted as MORE top speed, not necessarily HIGHER. (obviously not the primary point here, but still valid).
Also if you have no clutch, (in a perfect world with no belt slip) then you aren't wasting power making heat trying to get it to the wheel. Like trying to drink from a firehose.
consider trials bikes modifications on the other end of the spectrum. people often add rotating mass/flywheel weight to help with "more low end" power. If tuned with the other systems of the engine, it will help keep the bike in the low powerband longer, and with positive clutch engagement. so you are investing all of the power you can in the inertia bank to make things more tractable. (insert water drinking analogy here).
as far as french mopeds go, I think the "clutchless" setups I have seen were for simplicity. they still have a clutch, just in the form of belt slip.
I would consider racing without a clutch if the track was appropriate - which I am assuming would be higher corner speeds and longer straights. I've never ridden with one, but i'm certain you could tune/gear it for short stuff. typically geared for top speed, using clutch to get off the line. but ideally the variator has such a range that you spend as little time possible slipping the clutch to get out of corners, but still has full range to achieve top speed.
just my 3¢