Ask HN: Aggregating authentic user reviews across platforms?

I'm exploring the technical feasibility of building a tool that aggregates genuine user reviews about websites from various sources (social media, forums, review platforms, etc.). The core challenge: How do you programmatically collect and verify authentic user sentiment about a website while respecting rate limits, ToS, and privacy concerns? Technical questions I'm grappling with:

Data sources: Which platforms actually allow review scraping legally? Authentication: How to handle platforms that require login for review access? Rate limiting: Best practices for respectful data collection across multiple APIs? Spam detection: How to filter out fake reviews and bot-generated content? Real-time updates: Efficient ways to keep review data current without overwhelming source platforms?

Broader questions:

Has anyone built something similar? What were the biggest technical hurdles? Are there existing APIs or datasets that make this more feasible? What legal/ethical considerations am I missing?

Currently researching this space and would love to hear from anyone who's tackled similar challenges in review aggregation, web scraping at scale, or sentiment analysis. Any insights on the technical architecture or cautionary tales would be incredibly valuable!