Some researchers at Imperial College London recently published a paper in which they analyze EOS, Ripple and Tezos, finding some controversial conclusions. Dan Larimer in his latest blog post wanted to comment about the analysis on EOS.
The analysis of the paper is focused on the throughput of the three blockchains, concluding that although they are able to handle a high level of transactions, the purpose for which they were created has not yet been fully realized. In the case of EOS they find that 95% of the transactions were triggered by the airdrop of the EIDOS tokens, and this mechanism seems very similar to that of a DDOS attack.
In the blog post Dan Larimer criticizes their definition of throughput, as the authors of the paper redefine it as "only transactions they think have value", and not as any transaction. Indeed, the EOS mainnet as a public blockchain, has shown that it can handle 3000+ Transfers for Second, but the quality of these transactions must be determined by users, or rather by the market.
At the moment it seems that the EIDOS airdrop is the one that most stimulates the use of the chain, and therefore dominates the throughput. However in the future a new dApp might find another use-case and take over EIDOS. In that case EOS will have already shown that it can maintain a high level of transactions per second.
"All of that said, even with the authors poor definition of throughput, EOS is processing 2x the transactions throughput as Ethereum is even capable of if we assume no Ethereum transactions are spam or otherwise “valueless”."