Please make your identity more clear, e.g. what is your line of work, where you work, your experience etc. Like "Chris Hill" has done, and with his comment 22-Feb-16 I shall oblige with a reply. Unfortunately, from your "reputation" it appears you are spamming other authors work and promoting your self-interest manuscripts.
Probably article like this should be published in a peer-reviewed math journal.
Anyway, you should clearly state the asymptotic complexity of your P(n) function. Without it your objectives are extremely unclear.
Also, the discussion on 2 being "really" prime number is a gibberish. Number 2 satisfies the definition of a prime number. It "really, really" satisfies the definition of a prime number. End of story. It doesn't really matter if it fits into your 4-2 prime model.
The code as it is ready for 4-core processor computation. Just start-off each void Px(), x = 1..4 in a separate thread. Because you don't assign memory, you are not limited by how large N is. But don't forget, the code will run much faster if you really understand the structure of the prime number series members. For example, try setting RowJ1 values to TRUE and FALSE and observe the time difference in thread one (P1).
Please ignore the other user vote 1 feedback and enjoy the work presented here.