The scrape you’ve provided is essentially a JavaScript/cookie wall — Product Hunt served a bot-block page instead of the actual launch content. There’s no product name, company, pricing, description, or even a single URL in what came through, which means every fact the essay would need (who launched what, what it does, how it compares to incumbents, what the claim numbers are) is absent. I can’t invent a product or attribute made-up features and prices to a real Product Hunt listing, and the rules you set explicitly forbid fabricating facts the source doesn’t contain.
So per your own fallback instruction — “If the source is mostly chrome with no real content, return a single empty paragraph — don’t pad” — here is the correct output:
