Picturing Ideas: Timo Maas + the vision thing

Can you picture a vid for Timo Maas' Pictures?

A solid idea is central to a successful video project, to get momentum going, to make it possible to have a meaningful sense of target audience and how your media language choices will pander to them, to simply hold together the whole project.
That doesn't mean it has to be complex, or take agonised weeks to formulate.
It also doesn't mean you can't radically shift your idea once you start practice shoots and testing the idea out, not to mention the scope of post-production (editing) to transform an idea into a new shape and direction.
I was thinking about this on the way into work this morning, listening to the following track, Pictures by Timo Maas ... have a go at listening + brainstorming ideas for yourself before you read the text below the embedded vid... (the lyrics can always help this process!)

(you can also listen on soundcloud)
Here's the notion that hit me ...

From 0-0:28 (instrumental) I'm picturing one instrument at a time lit up in a darkened room, using abstracted framing in the style of one of the Pixies' iconic videos
An abrupt cut to one of the protagonists once the lyrics kick in at 0:28 - actually, start with cam mounted on car dashboard getting footage of the road, leave the driver as a narrative enigma...
When we get to the lyric about 'getting in my car', introduce a second character - I'd be tempted to apply Judith Butler's notions of gender as performativity and either have a masculine-signified (suit, short hair for example) female or the reverse (male in non-camp drag)
I'd eventually, after several teasing shots, reveal the driver ... and add a twist by going on to reveal a second character in the back seat. I'd clothe him similarly to the driver, as I want to signify a split personality - same brands/sizes, so perhaps some temporary tats or jewellry to signify some degree of binary opposition as well as the sameness
He is going to separate from the driver, and we'd signify him as controlling the 'other' half, introducing paranoic, perhaps mildly technophobic, themes not uncommon in techno music (ironically)
Cross-cut at points with the performance set-up, with some layering of this too
The darker character is the one who lip-synchs (tho' then again it might be interesting to cut between the 2 lip-synching)
We build up a stalking scenario, using photo boards (need to take LOTS of images at different times + places and print these off - we could very gradually a whole wall of these throughout the vid)
There is of course for a twist ending by the signified victim turning the tables, but I hadn't really thought too much about the ending, a critical part of the vid - would I really want narrative closure?

Feel free to post any thoughts/your own ideas below...