I'm going to go ahead with the update this Sunday morning, 17 April at around 6am.
The Japanese announcement has had the corrections applied, so it should be more readable. Unfortunately the person checking it had some trouble with the work, as I mentioned before, but I think it's basically OK.
There are a few small jobs to complete before it will be 100% ready, but I think it's more or less in good shape now, so I will just finish the remaining jobs, which are to do with logging of results etc.
I think the shape-only matching can probably be unified into the same binary as the stroke-order independent thing, but I'll hold off that change for now.