I'd steer clear of trying to do full character creation in a Con-game, you just won't have time to do that properly with any guarantee of the group being suitable for the scenario, or of having the time to finish that scenario.
The way I see it you have three options - I'm assuming that you're not planning on doing city creation as part of this.
Option 1 is to use the on the fly character creation. Supply everyone with the list of skills and trappings, a pre-defined skill slot package (so they know how many skill slots they have per skill level) and their apex skill, any musts for the template, a high concept and trouble aspect. Then just play, and they can fill in the blanks as necessary. I'd be inclined to make sure everyone has easy access to the stunts and powers list (probably very brief descriptions of each typed up on a few sheets of paper rather than out of the book - you want them to be able to scan through rather than having to wade through the book descriptions).
Option 2 is to create complete pre-gens. This has the advantage that you know the characters are appropriate, you can control what items etc they have, and you can tailor the scenario to best fit those characters without having to worry about tweaking it on the fly.
Option 3 is a combination of the two. Give the characters a High Concept, Trouble, Rising Conflict aspect and First Story aspect. Tie some characters together with a combination of these, but don't tie them all together. Leave the Guest Star slots blank. Give them three or four of their highest skills (and let them know how many slots they have for each skill level), and any must have powers, plus give each one a suggestion of stunts they may wish to take. At the start of the session give them ten or fifteen minutes to chat and come up with their final two aspects and the stories to fill out those guest star slots - give them the summary structure that the rulebook suggests for these stories and tell them they have to stick to that, tell them all to come up with a convention for the story titles (whether it's tarot cards, song names or whatever). Be strict about that time limit - if they can't agree then they'll have to come up with the last two aspects on the fly if necessary. This last one gives them a taster of the co-operative part of character creation, but limits the amount of time it will take.
Personally I'd go with option 2 or 3 - unless you're running the sort of scenario that forces diverse characters with completely separate interests to work together to survive. If that's the sort of scenario you're running then the only real thing to bear in mind is the time factor - which still means that you're best to do a large chunk of character creation before you get to the table. If the scenario requires any real degree of voluntary co-operation then I'd stick with 2 or 3 because they give you the most control over the characters that the players end up with.