Taking that model, let's make some assumptions and pul some numbers out of the air:
1. The TV companies would only agree if it is an exclusive service that doesn't take too many viewers away from broadcast, so the service would have to sit at a high price point.
2. That means that the only sensible way to do it would be to create an involved, high quality, ad free subscription product for committed viewers
3. Let's assume you could get the global web rights for £50million p.a. (I think this is on the low side,but let's say you could)
4. You could definitely put the product and infrastructure together for £10million, and add £5million p.a. for staff/maintenance/upgrades/bandwidth/backups/whatever
5. Let's budget £5million for content creation (interviews, insight pieces, etc), production and whatever else you feel like making (again, might be on the low side, but you could do it)
6. Looking at a 5 year contract, that means you would need a budget of 50*5+10+5*5+5*5+10% contingency = £352million
7. Let's assume a potential global audience of 1 million committed fans who can afford to pay a subscription (I think this is probably on the high side, but maybe)
8. Let's assume we can sell a new product on annual subscription rather than messing about with race by race or freemium
9. Let's assume we have a sh!t hot marketing team who get 25% penetration in y1, rising by 25% (net) p.a. = Y1 250,000, Y2 312,500 ,Y3 390,625 , Y4 488,281, Y5 610352
That gives total subscriptions over 5 years of 2,007,758. Let's call it 2 million for ease.
Total costs are £352million. Let's call it £350million for ease.
350/2=175
So, if you could get it all done at that price and get 2 million subscription sales, then you could sell it at £175 p.a. and break even over 5 years.
We've got a good cross section of the target audience here, so let's ask the question:
Who'd be willing to pay £175 p.a. to access the FOM coverage of all sessions (inc tests) and decent interviews and insight pieces?