The changes in the numbers of congregations for Mozambique (238 added) and Nigeria (496 fewer) are not consistent with the changes in the number of 'publishers'.
Mozambique did report a 7% increase, which was 4,887 new publishers since the previous year, which might initially seem to straightforwardly explain the increase in the number of congregations. However, this would average out to the new congregations having only 35 publishers each, though the average number of publishers per congregation in Mozambique is 48. So it appears that there are a significant number of considerably smaller groups that are being classified as separate congregations.
Conversely, Nigeria reported an 8% increase, with 27,268 additional publishers which isn't directly consistent with a reduction in the number of congregations (and certainly not such a significant drop), so it seems that many congregations have been merged there, perhaps combining previously small groups. It may also be that people have had to move away from parts of the country where there is significant opposition to Christianity generally.
In both cases, it is possible that at least some of the reported increases in publishers is partially due to people without reliable Internet access not being able to report activity for alternative preaching during the pandemic.
There could be something else going on regarding the congregation numbers in these two countries too, such as being in different phases of real estate Kingdom Hall building vs. selling?