test: add failing tests for Madison walkthrough findings

Six skipped tests, one per issue opened from the Madison walkthrough audit
(docs/walkthroughs/FINDINGS.md in cog_explorer). Each asserts the desired
behaviour, so it fails today and goes green when the fix lands; each is
guarded by a single skip() naming its issue and finding IDs, so the suite
stays green and activating a test is a one-line deletion.

  test-expenditure-concepts.R             #11  F-012, F-017, F-018
  test-revenue-concept-insurance-trust.R  #12  F-014
  test-coverage-disclosure.R              #13  F-020, F-023
  test-peer-summary-scope.R               #14  F-021
  test-amount-units-documented.R          #15  F-004
  test-gov-search-literal-match.R         #16  F-025

helper-walkthrough-raw.R reads the corpus's long parquet partitions directly,
bypassing uscogdata's SQL views. Every expected amount comes from there rather
than from the verb under test - verifying an absence through the filter that
creates it proves nothing, which was the most common defect in the audit itself.

Verified: with the skips removed all six fail (or error) against the bundled
fixture; with them in place the full suite is 576 pass / 0 fail / 6 skip.

tests/testthat/helper-walkthrough-raw.R

@@ -0,0 +1,44 @@
+# Helper for the Madison-walkthrough finding tests (uscogdata #11-#16).
+#
+# Those tests all assert something about what a `cog_*` verb includes or
+# excludes. The expected amounts must therefore come from the RAW corpus, never
+# from the verb under test: verifying an absence through the filter that creates
+# it proves nothing. `wt_raw_*()` opens its own DuckDB connection straight onto
+# the corpus's `long` parquet partitions, bypassing uscogdata's SQL views (and
+# therefore its `flow_prefixes` filtering) entirely.
+
+wt_corpus_glob <- function() {
+  url <- Sys.getenv("USCOGDATA_URL")
+  if (!nzchar(url)) testthat::skip("USCOGDATA_URL is not set")
+  paste0(sub("/$", "", url), "/data/long/**/*.parquet")
+}
+
+wt_raw_query <- function(sql) {
+  con <- DBI::dbConnect(duckdb::duckdb())
+  on.exit(DBI::dbDisconnect(con, shutdown = TRUE), add = TRUE)
+  DBI::dbGetQuery(con, sql)
+}
+
+# Sum of `amt` (in $1,000s, as the corpus stores it) for one government-year,
+# restricted either to an explicit set of item codes or to a set of first-letter
+# prefixes. Aggregate rows are excluded, matching every published verb.
+wt_raw_amt <- function(govid, year, codes = NULL, prefixes = NULL) {
+  stopifnot(xor(is.null(codes), is.null(prefixes)))
+  filter_sql <- if (!is.null(codes)) {
+    paste0("item_code IN (", paste0("'", codes, "'", collapse = ", "), ")")
+  } else {
+    paste0("LEFT(item_code, 1) IN (", paste0("'", prefixes, "'", collapse = ", "), ")")
+  }
+  out <- wt_raw_query(paste0(
+    "SELECT COALESCE(SUM(amt), 0) AS amt FROM read_parquet('", wt_corpus_glob(), "') ",
+    "WHERE canonical_govid = '", govid, "' AND year = ", year,
+    " AND NOT is_aggregate AND ", filter_sql
+  ))
+  out$amt[[1]]
+}
+
+# The item codes a verb reports having summed, flattened out of the
+# comma-separated `codes_included` column.
+wt_codes_included <- function(df) {
+  sort(unique(trimws(unlist(strsplit(stats::na.omit(df$codes_included), ",")))))
+}