Your attack method is very interesting. But there are two problems.
(1) You can obtain the IP addresses of all the nodes (see this topic), but you still cannot establish connections with all of them. Every node will have a threshold for connecting to other nodes.
(2) The proposer proposed that after different blocks are proposed in one slot, the validator nodes that receive different blocks will be cross-compared. In this case, the proposer will be slashed (i.e., LMD GHOST consensus). [see ETH2 book link]
My experience fully supports this point. I submitted my research to the EF twice: the first time, they told me it was already known, but I received a bounty from another vendor; the second time, an EF member outright plagiarized my work and proposed it as an EIP. I found that the plagiarist has been with the EF for a long time and profits from the Protocol Guild. He proposes many foolish EIPs, yet he remains a core developer in the community because he’s a sycophant of VK—you can find him in every VK post, quoting VK frequently in his posts. So, if you want funding from the EF, start memorizing ‘VK quotes’—it’ll greatly improve your chances.