You assert this as if this is how a judge would see it (or as if it is well tested). That seems highly unlikely to me.
It seems to be very easy to make the legal argument that because github knows which forks are copies and which aren't, they have actual knowledge of the ones that are still copies.
