If you want to make it chronologically correct, then you'll have a darn big challenge... we all need to record a clock, the same clock, in our recordings...
Oh well, maybe I am overthinking this again? For sure.... I am!
But, indeed, Sweper is correct, we need a track, complete.
Two people should record it, live. One to use, one in backup. Preferably, the host, and another one 'close to the host'
Next, view the track multiple times, and view all clients' contributions multiple times, then write a story board, of what to put when and where.
Run the track, changing the views at the correct times, (according to the story board) and making a base movie of it.
Next, insert all selected scenes from the clients' recordings into the base movie.
Put in subtitles, and there you go.
Simples.
Oh, I forgot one little detail... sounds. Well, to have the sounds correctly in the movie, all timed up nicely, and volume wise balanced.... piece of cake!
So, who of us has that much time on his hands?
Saying; it's is a good idea, I do like it, but do please realise that for a good movie of an hour playing time, one needs a multiple of that time to create the product....