Register weighted-microdata rules: weights aren't values #3937

Problem

A weighted MicroSeries or MicroDataFrame weights every .sum(), .mean() and .count(). Aggregating a weight column through one weights it by itself, and returns the sum of squared weights.

That happened in the Simulation API (PolicyEngine/policyengine-sim-api#766, fixed by PolicyEngine/policyengine-sim-api#767). budget.households was about 10^4 times too high on every run, and its test missed it because it used plain pd.DataFrames.

Once the unused economy comparison module is removed, nothing in the Public API handles microdata. Nothing stops new code from making the same mistake, though.

Proposal

Phase 1 step 1.9. Register the rule that weights aren't values, plus AST guards for the shapes a source check can recognise, in rules/registry.toml:

• api.microdata.weights-not-values (doc-only): weights, values, labels and counts each allow different operations.
• api.microdata.weights-not-aggregated-as-values: a weight is never the value of a weighted MicroSeries, the variable of a weighted aggregator, or frame[<weight>] followed by a weighted reduction.
• api.microdata.weight-stripping-confined: raw values (np.array, .values, .to_numpy() and so on) are read from weighted microdata only in a designated module. None is designated yet.
• api.microdata.unweighted-series-are-labels: a MicroSeries built without weights= is only a grouping or filter key.

The aggregation rule's proof is the real pre-#767 Simulation API line. With no current code in scope, all three guarded rules start enforced, with no findings.