I will try to give some genuine advice.
First off, get rid of the difficult house-wife attitude. You are an entrepreneur. The one common element between all of the successful entrepreneurs I have met before they were successful is that they exude confidence. You are not up against the city, you are tasked with informing the city and large agencies.
Second, being large agencies, they want proven test cases. Spending $1M on your solution has far more risk than $10M on a solution which has been used by other transit agencies. Why? Politically, you are asking them to extend all of their political capital on an unproven technology; and they really only get to push 1 or 2 special projects per election.
Third, you need to remove the major stumbling blocks.
A) Pick a specific mechanism. Doors, bars, guides, etc. Stop saying anything will work and pick a specific implementation to sell. When you leave the implementation so wide, they will feel they are designing it for themselves and don't believe in the project in the first place.
B) Advisory Committee on Accessible Transportation (ACAT) is a HUGE stumbling block with the ability to dictate damn near anything for a new project. Get them to sign off on your specific implementation.
C) Get a fire chief to sign off on a specific implementation as meeting fire code.
D) Timings. You need to prove it will reduce dwell time. Get 30 volunteers, including one person in a wheelchair AND a vision impaired person (Rocket Riders has/had both as members -- ACAT will help here) and do the following tests for your specific setup and the current setup (standing to left/right of doors):
- 0 off, 30 on
- 10 off, 20 on
- 20 off, 10 on
- 30 off, 0 on
- 25 on and 5 trying to get on but only 2 manage to squeeze in.
- 15 getting off but the exit gate is blocked or is broken
- 15 getting on but the entry gate is blocked or broken.
- Medical services required on train. 28 getting off, and a crash cart getting on.
- Fire/Smoke. All 30 people getting off and they're not friendly about it (pushing, shoving, etc.)
Show that these are improved or explain in detail why the situations which did not improve are not a problem. Note,
Fourth, handle platform crowding. The stations with the biggest crowd problems are King, Queen, Dundas, and Bloor. The first 3 have rather tight platforms and you are taking away several feet of standing space. How does this impact capacity. Have your 30 volunteers show they can still wait in this reduced space compared to today.
Actually go to King during Rush Hour and take rough measurements of how many people there are per square meter, make your reduction for "flow" control, and show it continues to work.
Finally once you have all of the above fleshed out and functional you need to do a real trial with a real train.
Luckily Lower Bay station and trains are rentable for a pretty modest fee (all things considered). For less than $10,000 you can do a real trial with a real train arriving packed with 100 volunteers onboard a single car and have 100 others waiting on the platform. Run through the exact same scenarios as listed above with your real situation. Take video evidence.
Now you are ready to market a unproven solution with minimal evidence of functionality to a transit agency along with your 10% cut, or $100k per station.
As someone who has gone from idea to profitable implementation with a few different products; the idea is the easy part. Making people want to buy it (proving efficiency/safety in this case) is the hard part and you need a ton of evidence to make it something the public will swallow without lots of negative press.
I'm friends with a couple of the people who build the OneStop display boards (hardware and software) and did the installation. You will need to do all of the above as a MINIMUM.
In actual probability, you will need actual installations for a few years somewhere other than the TTC first. A 5% increase in efficiency at Bloor station alone is worth $50M per year in operating costs/capital; likewise, a 5% decrease in efficiency will cost that much or more. A one time charge of $10M is pretty damn cheap when you compare against that -- you need to guarantee results without adding "or you get your money back".