Hmmm... this is a surprise to hear actually. I applaud your description of the scalability issues; the surprise is in the analysis. Although I realize this is the first announcement for a new design concept and you are excited about it, there are several concerns I have based on reading this.

First, how does your analysis factor hardware performance increases over the next 10 years? Last year Intel announced a dramatic new memory fabrication technology (3D XPoint™) that will have a huge impact on hardware performance and architecture when it becomes available, and I suspect we'll be seeing this surface in the next year or 2.

You certainly don't have to convince me how important replay time is, as that was a killer in BitShares 0.x.x. However, I'm surprised by the numbers here. I thought you had plenty of room to scale for the future built in based on early testing of graphene on BitShares 2.0 last year. I recognize Steemit is different, and you specifically mention voting in your analysis. Is that what renders any comparison between BitShares 2.0 & Steem unreasonable? Are there other differences between Steem & BitShares besides voting that make scaling Steem so much more difficult than BitShares?

Second, I'm concerned about how this announcement will impact Steemit. I'd hate to see a repeat of the issues that surfaced from the BitShares community when graphene was announced last year, and further the impact implementing your new design will affect existing Steemit apps and efforts which are far more numerous in this community.

@dan, you are an amazing tech wizard and I'm sure your analysis is sound, and I look forward to an updated roadmap and additional details. Despite that I feel uneasy as this announcement seems so similar to last June.