It's interesting to see how so many people are still overlooking EOS. Now that a simple kitten game (one of the only projects actually completed/up and running on Ethereum) is making Ethereum crawl, it should be obvious that it's scalability issues are a much bigger problem than people realize.
Regardless of its scalability issues, the fact that it requires the use of "gas" makes it a flawed system for mass adoption. There will never be a successful social network similar to Facebook, Twitter or even Steemit on Ethereum, because the "gas" model would require everyone to pay a fee to send messages and make posts.
Ethereum:
Transactions per second: 30
Block Times: 20 Seconds
Fees: Gas
Coding Language: Solidity
EOS:
Transactions per second: Millions
Block Times: Half of a second
Fees: None
Coding language: WebAssembly and several other popular languages
The hundreds of ICOs that already launched have been planning on deploying their dapps on Ethereum, but these projects will never see completion as a result of Ethereum's flawed architecture. Interestingly, Dan Larimer, the creator of Steemit, BitShares and EOS foresaw this way back in 2014.
I believe we will see a mass migration of Ethereum based projects moving to the EOS blockchain in 2018. Each time this happens, news of the migrations will spread, and we could very well see a consistent decline in the price of Ethereum, and a steep incline in the price of EOS. With its one billion coin supply and all of the new money flooding into the crypto market, I see no reason why EOS won't be $50/token in 2018 and much more in the years to follow... Good luck Ethereum!