Two days work done... getting there.
My game is deliberately simple and silly in the Flappy Bird / Crossy Road sense.
Why? So I have an opportunity to play with the browser/DOM as a game engine, controlled via PyScript, without having to worry about complicated game context and in between "real work" demands on my time.
I think I'm nearly there.
Today I managed to check off most of my outstanding items on my to-do list:
- There's now a scoring system!
 - The aquatic life moves and animates.
 - BUBBLES!
 - It's a responsive web app, so displays well on mobile browsers.
 - I've adjusted various settings so as the score increases the game becomes faster / more challenging.
 
Some screenies:

^^^^ A responsive layout of the game!

^^^^ There are a selection of different animated backgrounds.
Of course, you can see it in all its glory via this YouTube video: https://www.youtube.com/watch?v=-nni8GHdS9M
Here's what remains:
- Add a countdown clock, so you have to get as many points within the time limit.
 - When on mobile, use the browser based accelerometer API to control the submarine's depth (i.e. tilt the phone to steer up/down).
 - Play test with my kids, folks at tomorrow morning's gym class (they were enthusiastic to learn I was making a game for PyWeek), and my colleagues at tomorrow's PyWeek community call (on discord).
 - Polish / refactor the code. Some of it is a bit slap-dash as I'm spiking. Nevertheless it's still < 360 lines of Python (including comments).
 
I think I should be finished tomorrow, with plenty of time to spare. 
All in all I'm having lots of fun making a goofy game while exploring the game-making potential of the browser/DOM. Post PyWeek, I'd like to explore how we could perhaps use something sensible like the <canvas> element to make a "real" game engine in the style of PyGameZero or Wasabi2D.
Onwards!