You may need to share more information especially on the rates using 2.21. To see the margin difference.
Could it be that in 2.21 there are alot of new facilities what were being created moving forward that caused change in reporting rates for the previous periods when the facilities were not expected to report?
With 2.24, reporting rates are based on when the facility was created i.e the opening date of the facility
Expected reports based on org unit opening and closed date: For reporting rates in analytics, the number of expected reports is now calculated based on whether the organisation units assigned to the data set were open or not for the aggregation period. As an example, if you ask for reporting rates for a district for June, and one of the facilities within the district was not open until August, that facility will not be included in the number of expected reports.