minor docs site content tweaks

Author: sdcondon

src/SCFirstOrderLogic.Documentation/Razor/Pages/GettingStarted.razor.md

@@ -238,7 +238,7 @@ using SCFirstOrderLogic.Inference.Basic.Resolution;
 var clauseStore = new FeatureVectorIndexClauseStore<MaxDepthFeature>(
     MaxDepthFeature.MakeFeatureVector,
     new ClauseStoreFVIListNode<MaxDepthFeature>(MaxDepthFeature.MakeFeatureComparer()));
-// ..or var clauseStore = new HashSetClauseSetore(); for a simpler, less powerful clause store.
+// ..or var clauseStore = new HashSetClauseStore(); for a simpler, less powerful clause store.
 
 var kb = new ResolutionKnowledgeBase(new LinearResolutionStrategy(
     clauseStore,

src/SCFirstOrderLogic.Documentation/Razor/Pages/lab/FeatureVectorIndexVisualiser.razor

@@ -28,7 +28,7 @@
             </p>
             <p>
                 Feature vector indices (in this context, at least) are an indexing structure for CNF clauses that enable quick lookup of stored clauses that subsume or are subsumed by a query term.
-                The general idea is that for each stored clause and query clause, we calculate a set of "features" (essentially, metrics) that will always have a greater or equal value for subsumed clauses.
+                The general idea is that for each stored clause and query clause, we calculate a set of "features" (a feature being some non-negative integer value computed from the clause) that will always have a greater or equal value for subsumed clauses.
             </p>
             <p>
                 Note that:

src/SCFirstOrderLogic/ClauseIndexing/AsyncFeatureVectorIndex{TFeature,TValue}.cs

@@ -24,7 +24,7 @@ namespace SCFirstOrderLogic.ClauseIndexing;
 // TODO-BREAKING: Support for cancellation.
 // TODO-PERFORMANCE: at least on the read side, consider processing nodes in parallel.
 // what are some best practices here (esp re consumers/node implementers being able to control DoP)?
-// e.g allow consumers to pass a scheduler? allow nodes to specify a scheduler? or just expect caller to manage via appropriate context?
+// e.g allow consumers to pass a scheduler? allow nodes to specify a scheduler? or just expect caller to manage via sync context?
 public class AsyncFeatureVectorIndex<TFeature, TValue> : IAsyncEnumerable<KeyValuePair<CNFClause, TValue>>
     where TFeature : notnull
 {