There is a large miscount of provisional items, which will cause vetters to miss items they need to confirm (or fix). (This is on smoketest, but I have no reason to think it will be better on production.)
Example: https://cldr-smoke.unicode.org/smoketest/v#/de/Area/9a46cd100a26c50
This item is marked as provisional, and needs to be reviewed, but doesn't show up on the Dashboard. https://cldr-smoke.unicode.org/smoketest/v#r_vetting_json/de//
There are a hundreds of items like it.
—
Background
I just ran Priority Items (on smoke test). It shows 43 provisional items
Level | Locale | Total | 0 | 10 | 1,448 | 2 | 43,161 | 40,569 | 68,558 | 26,702 | 28,150 |
modern | German | de₍_Latn₎₍_DE₎ | 0 | 2 | 43 | 2 | 442 | 609 | 407 | 242 | 285 |
https://cldr-smoke.unicode.org/smoketest/v#vsummary///
When I look at the right-side bar in the German Dashboard, I see Provisional (45/45). That is close, but not the same (when it should be).
https://cldr-smoke.unicode.org/smoketest/v#r_vetting_json/de//
However, when I look at the XML file, the new seeded data has many more provisional, such as the following 6. In fact, there are 649 total in de.xml.
When I look at the right page for that on smoketest
https://cldr-smoke.unicode.org/smoketest/v#/de/Area/, it is indeed showing those 6 items as provisional. (X's in the "A" column instead of checks).
So for some reason they are being miscounted in both the code that does priority items and the code that does the dashboard.