The problem of why ycombinator model has not worked in India is the same as why other ycombinator copycats in other parts of the world have failed too. (Besides TechStars, I can't think of very many successful ycombinators in other parts of the world - including Canada and Europe and Australia.)
The reason is: these copy cats have not cracked the code of attracting and accepting awesome startups. The success of ycombinator is mainly based on the success of its startup selection.
You need someone like Paul Graham to attract a lot of startups. You then need to be good at filtering out and select only the startups that show promise. And you then need to help these startups with huge demo days.
Who else has managed to crack this? No one. If you see the Indian and other ycombinator copy cats, you will still see such a high degree of crappy startup acceptance rate. Fix that, and you will have a successful ycombinator copy cat.
Because the market is global.