The current `PDFExport` / `GridPDFExport` pipeline relies on `drawDOM`, which traverses and renders the full DOM tree on the main browser thread before generating the PDF. This works for smaller datasets, but it becomes impractical for large Grid exports. In a reported real-world SPA scenario running behind CloudFront with no backend, exporting 10,000+ rows takes approximately 10 to 15 minutes, blocks the UI, and makes the built-in PDF export unusable for production workflows.