Not exactly - sometimes, the bug still happens in the current build, so the repro steps still work. If that doesn't work, we do keep the current public build available so that we can look at the issue you report.
We are particularly interested in how the UI will work when we have a bigger data set. for example, right now, searching for a bug before you submit doesn't seem onerous because the result set is pretty small and easy to scan through. What happens when
the result set has 50 issues, or more?
I do like your suggestion for a watcher/crawler app. We've talked about doing something similar to pre-populate the version and config fields,but it didn't make the list for v1. but your comment goes beyond what we were thinking. you should submit it.