I find it hard to believe that you wouldn't lose any time you gain in loading/unloading to transit time inconsistency when you have to rely on wind. This is not even to mention the fact that these ships cannot make good use of big modern ports with cranes and the best logistical connections and that it would take 100+ of them with hundreds of crew to move as much as a single container ship can with 20 people. I could see them being useful in niche scenarios like cabotage within island/archipelago nations or shipping small loads on irregular schedules, but for anything else it is very hard to beat a container ship.

This is HN, we can use an analogy of TCP window size or UDP packet size in an underlying high latency medium and the receiver and sender have very high processing costs. So perhaps solving the unloading and loading logistics is also worth optimizing for like we did almost for free in computing space? But because we don't have ballpark numbers of each terribly well we're going to have some difficulties with a valid non abstract system design discussion.