While procrastinating actually writing the prompts, I've been pondering how likely prompts actually are with a single roll of the dice. I think a large swath of the prompts just aren't reachable with a single roll of 10d6. That might be fine, since you can still use fate to get to them, but as a limited resource that means a lot of the prompts I write just are going to be unlikely to see use. Prompt 12 in Concept / Location 3 is unreachable unless you can spend more than 6 points of fate for example. 

I imagine handling this by just rerolling each act after determining  the location / concept for the scene. Anyone else thinking about this? 

I just read through the SRD and this was my first thought. Since it basically involves rolling a pool of dice to generate a Concept and further prompts, you will be creating a bell curve of prompts which are reached more often than others. I applaud your break down because I hadn't gone that far in depth!

Apologies for not responding sooner to this. Thank you so much for putting in the work and calculations to generate this graph and you are absolutely correct! This is on my list of things to update on the ToL document.

Submitted my prototype for the jam in case people are interested. Here's what I'm thinking to keep the game roughly following the original SRD, but to try and make sure all the prompts have a chance to show:

1. Reroll 3d6 for each Act (I'm calling them scenes) - this keeps the probability curve the same for prompts within each act, making them all at least in theory reachable, but moving away from the cool concept of "everything from a single roll".

2. Swinging the initial roll further toward extremes by adding/subtracting more from the thread of destiny each play through. Given #1, this mostly just impacts the closing prompt (I decided to use them as openers - "episode arcs") and the acts.