Benchmarking TTL list caching: a 100,000-request look at Appwrite at scale

We ran 100,000 listRows calls per phase against a 10,000-row table, with and without TTL caching, to measure the latency gains. The results were more decisive than we expected.