FIX: Implement smart caching to avoid CoinGecko API rate limits

[mb43]

Problem:

• Fetching every second caused 429 rate limit errors from CoinGecko
• Free tier API has strict rate limits

Solution - Smart Caching System:

• Add api_cache dict with configurable TTL per data source:

  • Price data: 60 second cache (low volatility)
  • Network stats: 5 second cache (more dynamic)
  • Block reward: 30 second cache (rarely changes)

• Cache logic:

  1. Check if cache is valid (within TTL)
  2. Return cached value if valid (mark as "cached")
  3. If expired, fetch fresh data from API
  4. On API failure, fall back to expired cache
  5. Update status indicator accordingly

Features:

• is_cache_valid(): Check cache age vs TTL
• get_cached_value(): Retrieve from cache
• set_cache(): Store with timestamp
• Graceful degradation on API failures
• Visual feedback (orange=cached, green=live, red=failed)

Cache TTLs:

• Prices: 60s (avoid rate limits)
• Network: 5s (near real-time)
• Block reward: 30s (slow changing)

This allows 1-second UI updates while respecting API limits.