That's cute, but it's probably not going to work. The article's author is confused about how a 51% attack works. What BitPico seems to be planning is a "stress test" of extremely large transactions. They believe the following will result:
the network will start forking into multiple chains as the backlog in this manner impacts subsets of the network. Its a global blockchain stall that would eventually reorganize or recursively reorganize until time to infinite.
They plan to release a hard fork and they believe this will cause a multi-chain fork. In reality, the 5000 "attack nodes" are irrelevant in a hard fork because legacy nodes won't communicate with them.
And something tells me Bitmain won't let their baby be attacked so easily. If the majority of Bcash's hash rate is controlled by Bitmain or its proxies (and I believe that's true), they may be able to sidestep the attack very easily by censoring the attack transactions. The attack transactions will apparently be extremely large (
their example is an average transaction size of 0.6 MB). BitPico assumes that only rational mining incentive dictates which transactions are mined:
So no there is no motivation to mine small transactions over large transactions as it comes down to maximum profit; something complex blocks guarantee.
That's pretty absurd when a well-capitalized company like Bitmain is extremely invested in Bcash's success. I would also be very curious to see what funds BitPico has set aside to pay transaction fees. Attacks like this are not free, and their model depends on driving transaction fees up.