The answer is between "no" and "few".
This has always been a problem in RPGs. It is not that the rational and logical thing to do is to NOT surrender. In almost all cases, it is the rational and logical thing to do. But in this case, that's not the test.
This result is the same if, by slightly rephrasing the question, you substitute the word "flee" for the word "surrender". That action and that result also seems reasonable and fair when judged on a dispassionate basis. Indeed, usually more logical and rational than surrender ever would be, right?
Problem is, any encounter or scenario which posits as the "correct" response by the players (and make no mistake - we are talking about PLAYERS here, not "PCs") will be to surrender or to flee plays against the choices the overwhelming number of players will ACTUALLY make during a game.
The players are there to play and BE heroes. Not to run, not to surrender -- but to fight. Even if it is irrational, illogical or a foregone conclusion to the scenario - the overwhelming majority will not run, they will not surrender - they will choose to fight.
Accordingly, any adventures which posits that the characters "should" run or surrender is poor adventure design. GMs think this is acceptable because they see it in a book or a movie as a rational plot device. It makes "sense" to them that the honourable and "correct" choice in the circumstances it to flee or surrender.
Problem is, works of literature and film are works of literature and film. They are not RPGs. What works in one setting does not work in another.
This design is bad not because it is illogical or irrational -- nor because it sets up a scene which is not justified in the circumstances. It is poor adventure design because it is unheroic, is not fun to play, and is not what the players of the game are there to do. The players don't want to do this. You are making the game SUCK for your players when you "expect" them to make this choice. They will almost always resist it and will feel railroaded and pissed off.
The solution is simple: don't do it.
Don't do this in your games. Ever. It's a mistake - ALWAYS. If the players are going to be stupid and enter into a no-win scenario of their own accord, attack them, reduce them to unconsciousness and take them prisoner if that is what you want to have happen.
But don't offer up a fool's choice like "expecting" them to run or surrender as if that was ever a fair or reasonable expectation for you as GM to have. Games aren't real life and the overwhelming majority of players don't willingly choose that course of action. Don't delude yourself into thinking you are being a "good designer" when you offer this path as the way out. You aren't: you are running a crappy game.
I've done it too. I have done this in the past and allowed in-game reason and logic to dominate my design as to what a PC "should" do, and allowed this to prevail over my actual meta-game knowledge of what the players WILL do.
It's not the players' fault, it was mine. I SUCKED.
Pro designers sometimes make this mistake too. Don't make that same mistake just because they had an editor who didn't say "no" -- when he or she should have.
This is a difficult issue to discuss. But sometimes, logic and common sense and literary devices must give way to the operation of logic at a meta-game level.
Don't do this in your games. Just don't.