docs: improve product positioning, add FAQ section to README

[EricNGOntos]

Summary

This PR enhances the project README with two key improvements:

1. Product Positioning & Slogan

• Added a prominent slogan at the top: "We're not building the next document parser — we're building document memory infrastructure that agents can actually consume."
• Expanded the intro to clarify Knowhere's relationship with open-source parsers (MinerU, Docling, Marker): we stand on their shoulders and actively improve upon them with memory-oriented parsing optimizations and agentic parsing.
• Rewrote the one-liner to emphasize persistent, navigable memory for AI agents.

2. FAQ Section

Added a "Frequently Asked Questions" module addressing the most common community questions:

• Is MinerU required? — No, any doc-to-MD tool works; MinerU is just our current best-performing default for PDF/PPT.
• LLM/VLM dependencies? — DeepSeek for text/hierarchy, Qwen-VL for vision tasks, fully model-agnostic via env vars.
• Agentic vs Traditional RAG? — Explains hierarchical navigation vs flat vector lookup.
• Multi-modal support? — Images/tables extracted, summarized via VLM, and explicitly linked to source text chunks.

3. Benchmark Expansion Note

Added a [!NOTE] callout indicating benchmarks are actively expanding beyond the current MinerU baseline.

All changes are documentation-only (README.md). No code changes.