Many of the test queries used in pre-graph-split wdqs won't be relevant on scholarly or (possibly) main. We should audit the existing test queries and come up with specific examples for each graph split type.
A query for testing wdqs-main:
select ?paper ?type { VALUES (?author) {(wd:Q1042470)} SERVICE wdsubgraph:scholarly_articles { ?paper wdt:P50 ?author ; wdt:P31 ?type } } LIMIT 10
A query for testing wdqs-scholarly
select ?paper ?type ?firstname ?lastname { VALUES (?author) {(wd:Q1042470)} ?paper wdt:P50 ?author ; wdt:P31 ?type SERVICE wdsubgraph:wikidata_main { ?author wdt:P735 ?firstname ; wdt:P734 ?lastname } } LIMIT 10