Clarify grouped calibration target values in explorer/table

[MaxGhenis]

The calibration target explorer/table can make grouped target families look like a single aggregate value. Elaine flagged the EITC child-count row, which showed a value around .5B next to grouped geography counts. That value is a child-count slice from the IRS SOI EITC target family, not the all-in EITC target.

For EITC, the checked-in TY2022 SOI shape has child-count amount buckets of roughly .63B (0 children), .18B (1 child), .25B (2 children), and .00B (3+ children). After scaling to the IRS TY2024 EITC Central domestic control, those become roughly .05B, .53B, .78B, and .22B. Showing one grouped value without its cell definition makes it easy to misread as the total EITC calibration target.

Improve the tool so grouped target rows show:

• whether the displayed value is a total, representative cell, min/median/max, or sum across cells;
• units separately for dollars vs counts;
• source year and target year;
• target concept, e.g. IRS tax-year EITC claims vs Treasury/CBO fiscal-year outlays;
• the metric/cell definition, e.g. EITC amount by child-count bucket vs tax-unit count by child-count bucket;
• geography/cell counts separately from target values.

This should prevent rows like "EITC child count" from being interpreted as a .5B all-in EITC total.