docs: document lru_cache async footgun in comparison guide

[27Bslash6]

Description

functools.lru_cache on async functions is a well-known footgun — it caches the coroutine object, not the awaited result. Second call raises RuntimeError: cannot reuse already awaited coroutine.

This is a genuine cachekit advantage that should be documented in the comparison guide.

Evidence

from functools import lru_cache

@lru_cache(maxsize=128)
async def fn(x):
    return x * 2

result1 = await fn(1)  # Works — returns 2
result2 = await fn(1)  # RuntimeError: cannot reuse already awaited coroutine

cachekit handles this correctly:

@cache(backend=None, ttl=300)
async def fn(x):
    return x * 2

result1 = await fn(1)  # Works — returns 2
result2 = await fn(1)  # Works — returns 2 (from cache)

Suggested Fix

Add an "Async Support" section to docs/comparison.md documenting:

• lru_cache: caches coroutine, not result (broken)
• cachetools: no async support
• aiocache: async-first (works)
• cachekit: works with both sync and async (auto-detected)

Found by: tests/competitive/test_head_to_head.py::TestAsyncSupport