>>2) The GROUP BY could have been fixed by ensuring that all fields INCLUDING inherent
>>"field function" values were properly sourced from the correct record.
>
>Jim, the whole point of that change is that THERE IS NO CORRECT RECORD when using GROUP BY. It's completely ambiguous.

Sorry, Garrett, but I don't understand what you mean here. Tamar provided a simple example and in that case the MIN() value came from a different record than the other value. In the case of a CNT(*) I would expect that the "correct" record would be one where ALL other non-field-function values in the SELECT are the same regardless of whether they are in the GROUP BY or not.