There is a slight flaw in your statistical approach to valuing...

There is a slight flaw in your statistical approach to valuing this company. The "coin flip" analogy and approach works out right in the long run, over a larger number of coin tosses. You cannot narrow this 50% expected outcome over the long run to a single event and then say that an outcome on this single even will be 50% and therefore you can halve the 100% estimated value and use that as a base case value on a one time event because it will never materialise as it is not an achievable outcome on a one time event based on the original parameters that you have set out - it simply won't occur. On this single event it will still be either 0% or 100%.

If you had 1,000 PAR's lined up all doing similar things and all doing that at the same time, you invested $100 in each of them, and they all had a 50% chance of 0% success and a 50% chance of 100% success, then yes the expected win rate on your 1,000 bets would be that 500 would be dead losers and you'd lose all your money and 500 would be terrific winners and you'd gain multiples.