Details matter on these platforms. Here are the important ones on this issue (copy pasted from @dana-edwards  recent post https://steemit.com/steem/@dana-edwards/its-a-very-bad-idea-to-power-down-or-sell-steem-right-now-1518490604)

It's actually much more safe, valuable, effective to build these on Steem than any general platform, eos or anything else. I wouldn't dare code these on a general platform and purport they would work very well nor that they could have the same economic benefits to the core token. Ultimately, the idea of general purpose blockchains will also mean meta Proof of Brain tokens are limited by bad bandwidth or transaction pricing and other economic pitfalls, and in the end, application-specific blockchains will over-win opportunities like layered Proof of Brain tokens, such as SMTs.

Check out these sections of the SMT whitepaper. It mentions Ethereum, and the drawbacks of eos will apply just the same:

What Makes SMTs Better Suited to Application-Specific Blockchains, such
as Steem, than Application-General Blockchains, such as Ethereum? 53
SMTs are Safer and More Cost Effective in Application-Specific Blockchain Environments
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
SMTs on Steem have Aligned Proof-of-Brain Incentives with the Core Token . 54
SMTs on Steem Have Transaction Pricing that Contributes to a Quality User
Experience . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
SMTs Benefit from a Blockchain that has Scaling Processes Programmed to a
Specialized Set of Applications . . . . . . . . . . . . . . . . . . . . . . . . 55
SMTs Benefit from a Blockchain with Content Management System (CMS)
Primitives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56