Why do the "expected" numbers include teachers in the "Calculate assessment phase", hence a mismatch between expected and calculated? The other phases only seem to include students. Is this the reason for the red cross in the Calculate Assessment phase?
In the Submission phase, the task "Allocate submissions" takes all potential submission authors into account. The "potential submission author" is any user enrolled in the course with the capability to submit their work. In default settings, only Students are considered as potential authors. In your course, there seem to be 92 students and all of them submitted their work to the workshop and all the submissions were allocated for assessment.
Now in the Grading evaluation phase, grades for assessment are to be calculated. Any user with the capability to assess allocated submissions is considered as potential reviewer and the assessment grade is expected for them. In your case, it seems that your have overridden the Teacher role's capabilities and allowed peer-assessment to them. Hence, there are some 7 users (teachers) who were expected to assess some submissions. Most probably no submission was allocated to them (?) so no "assessment grade" could be calculated for them - and that's why the task is marked as "failed".
Note that you do not need to give Teachers the capability to peer-assess submissions. Teachers can simply assess any submissions they want even without the capability. The capability "mod/workshop:peerassess" is mostly used by allocation methods so that they know what users can be assigned as the reviewers of submissions.
Excellent. Thanks for clarifying.