I have been looking at the fork graph and what it looks like is that connections are stalling. What I am seeing is that the seed node which has init delegates 12 to 100 is producing blocks one after another but the rest of the network ( vikram1, bdnobel, delegate-baozi, vikram11, vikram2) act like they never got the memo even as there are 0 transactions at this time.
This is almost certainly a network / communication issue potentially caused by socket inactivity and TCP/IP connections going stale.
This is good news: the forks that have occurred are not for differences in evaluation logic, but instead purely differences in timing / communication.
I am working on something that should address this and hope to have it out tonight.