GroupBy aggregation don't double count rows that have the same Id

[akraines]

I'm using the built in group by feature to aggregate rows by either sum or count.
However I hit an edge case that is not handled by the built in groupBy aggregate function. The issue is that I have a list of items coming from a database query. I only wont to count each row once when doing the aggregate - if the same Id appears twice I dont want it to be summed twice.
I managed a hacky solution by attaching a Set to the column object that is passed to the aggregateValue function and then removing it when the Aggregated cell is rendered - this way I "initialize" the Set whenever it starts aggregating a certain column and reset it after all the work is complete for the next run. So far so good. I works for the built in Sum aggregate function. I will need to specialise the Count function to actually iterate through the rows and ignore undefined/null (or some sentinel value depending on my requirements if undefined should be counted).
However...
The aggregatedGroupedRows function only uses the aggregateValue function for the outermost grouping [depth === 0], from then on it just uses the regular aggregate function [whatever was defined for that column]. The issue is that for deeper levels of groupings I have no way of controlling which value should be used for a given row - it just returns the value from row.values[column.id] and does not give me a chance to intercept it.

 const leafValues = leafRows.map(row => {
            **let columnValue = row.values[column.id]** 

Any suggestions how to solve this? Is my only option to fork the useGroupBy plugin and change its logic?