I am resubmitting this idea because it was archived prematurely in my opinion. Although search by secondary ID has been implemented, sorting serves a different purpose:
Our use case is that lecturers receive paper exams sorted by student ID number and grade them in the according order, on paper. They then go to enter the grades in the Gradebook. It would be much easier for them to quickly enter a high amount of grades if they could sort the students by ID (LoginID in our case), not name. Exams are anonymous so names are not a sorting criterion we could use in this case.