Update LTF.md

doc/LTF.md

@@ -4,6 +4,7 @@ An experiment inspired by JSON-LD's expanded and flattened forms, followed by qu
 * What if we detach identifiers from nodes and limit type declarations to a single one?
 * Would this help reduce processing complexity, improve speed, diagnostics, and analytics (i.e., adaptability), and could it enable faster canonicalization?
 * Would that impact the design of type and definition systems?
+* and What if we simplify JSON-LD context definitions by reducing their capabilities and expressivity to simple mappings? This approach could enable greater analytics and facilitate comparisons to identify differences in definitions and changes in meaning.
 
 ### Traits