🐌 Fix performance regression causing Purdue.io front-end timeouts

[haydenmc]

Recently the Purdue.io front-end started showing timeouts when browsing the Class list page for a particular Course.

Investigation revealed expensive SQL queries being generated, taking upwards of a minute to complete. Analysis revealed this section was contributing most of the cost of the query, requiring sequential evaluation of rows from disk:

SELECT t8."Id", t8."Email", t8."Name", t8."MeetingId", t8."InstructorId"
FROM (
	SELECT i."Id", i."Email", i."Name", m0."MeetingId", m0."InstructorId", ROW_NUMBER() OVER(PARTITION BY m0."MeetingId" ORDER BY i."Id") AS row
	FROM "MeetingInstructor" AS m0
	INNER JOIN "Instructors" AS i ON m0."InstructorId" = i."Id"
) AS t8
WHERE t8.row <= $7

Further investigation revealed that this was due to pagination logic being applied to the $expand operation. An issue is open with a similar repro here: OData/AspNetCoreOData#1041

The pagination logic was being applied due to the recent change #65. This PR reverts that change to fix the performance regression until a workaround is provided by the OData library.