This is an interesting concept, I have to say. The moment of wonder would be when I realized I am controlling the footage being watched, which in turn becomes a tape that changes every time it is viewed. The two invisible viewers' discussion also serve as a hint to us as what to do next. A clever idea, indeed.
However, the execution is not very elegant, which is my main issue -- It is confusing and difficult to make out the environment. Partially because of the lighting but mainly because of the art :
I don't know where the door is. From the direction the girl walks, and the hint of "hiding behind the chair before running for it", I get the impression it's on the lower right? Maybe? I dunno. After 6 repeating same sequences, I give up.
I also dare others to say they knew that purple-ish rectangle is a chair the first time they see it. (I thought it's a table) Why do you pick the same color for the floor? They just all get mixed up. I have to ramp up my brightness a lot to see the backrest. And the proportion is just wrong, it looks like a chair for giants.
Wait, the thing on the middle top is a shelf of some sorts for CDs? I thought it's some strange hole with a pole?WAT?
The nameless guy's walking animation also needs refinement. It is ok for the the girl because of her nature, but this will not work because he is human. It's very jarring.
All that being said, I still appreciate the idea as a whole. This has some potentials and I hope you put some more work into this.