To be clear, I wouldn’t suggest this approach for evaluating a complete rewrite. In a situation like that you have enough information to make a more precise judgement in terms of cost/risk/reward. The purpose of this analysis is to help me determine how much might be missing in building a *brand new* application, with requirements I don’t even know about yet. I agree that unread is surely not a big deal to write, and rollout probably wouldn’t be a huge issue either, although it does more than you might think.

None of these things should be considered “black marks” against the elixir ecosystem, which as far as I’m concerned is an absolutely fantastic ecosystem. It’s just new, and we want to measure what impact that newness might have on a real world project.