Simulate neuroglancer requests to create a benchmark

Write a benchmark that simulates requests to a tiamat pipeline similar to how a user would access it via a running neuroglancer instance. The benchmark should measure the response time for each request.