I really like the idea you have here. To my eye and it's an observation I seem to make every time I've looked at a shot where these kinds of backdrops are employed. The photographer knows it doesn't look totally lifelike and yet many, many people shoot them in reasonably sharp focus.
Now it might be a case that the backgrounds are small, in which case I'd be off in the distance with a long lens and a large-ish aperture. It's a lot easier to use a healing tool or similar to take the creases out of the background if the focus is softer. And does the background need to be sharp, if it carries enough detail to present the situation it is alluding to
Maybe shoot the same layout with a softbox behind the subject so it rim lights the subject and have it turned down enough to keep the background reasonably lit so the window works. Then play around with the key light,with the subject further forward until you get everything balanced.
Regards
Tim