Update performance numbers (#307)

Author: kylebarron

README.md

@@ -30,7 +30,7 @@ The simplest, highest-throughput [^1] Python interface to [S3][s3], [GCS][gcs],
 
 <!-- For Rust developers looking to add object_store support to their Python packages, refer to pyo3-object_store. -->
 
-[^1]: Benchmarking is currently ongoing, but [early results](https://github.com/geospatial-jeff/pyasyncio-benchmark/blob/2d8ee4df8c2d93463323394f09fd094ad0381122/test_results/cog_header_results.csv) indicate 9x higher throughput than aioboto3 and fsspec.
+[^1]: Benchmarking is ongoing, but preliminary results indicate roughly [9x higher throughput than fsspec](https://github.com/geospatial-jeff/pyasyncio-benchmark/blob/fe8f290cb3282dcc3bc96cae06ed5f90ad326eff/test_results/cog_header_results.csv) and [2.8x higher throughput than aioboto3](https://github.com/geospatial-jeff/pyasyncio-benchmark/blob/40e67509a248c5102a6b1608bcb9773295691213/test_results/20250218_results/ec2_m5/aggregated_results.csv) for many concurrent, small, get requests from an async context.
 
 ## Installation
 

docs/performance.md

@@ -34,11 +34,11 @@ Let's consider the areas where we expect improved, possibly-improved, and equal
 
 **Many-request throughput with the asynchronous API** is the primary place where we expect significantly improved performance. Especially when making many requests of relatively small files, we find that obstore can provide significantly higher throughput.
 
-For example, [preliminary results](https://github.com/geospatial-jeff/pyasyncio-benchmark/blob/fe8f290cb3282dcc3bc96cae06ed5f90ad326eff/test_results/cog_header_results.csv) indicate roughly 9x higher throughput than aioboto3 and fsspec. That specific benchmark considered fetching the first 16KB of a file many times.
+For example, preliminary results indicate roughly [9x higher throughput than fsspec](https://github.com/geospatial-jeff/pyasyncio-benchmark/blob/fe8f290cb3282dcc3bc96cae06ed5f90ad326eff/test_results/cog_header_results.csv) and [2.8x higher throughput than aioboto3](https://github.com/geospatial-jeff/pyasyncio-benchmark/blob/40e67509a248c5102a6b1608bcb9773295691213/test_results/20250218_results/ec2_m5/aggregated_results.csv). That specific benchmark considered fetching the first 16KB of a file many times from an async context.
 
 ## Possibly improved performance
 
-**Using the synchronous API**. We haven't benchmarked the synchronous API However, we do release the Python [GIL](https://en.wikipedia.org/wiki/Global_interpreter_lock) for all synchronous operations, so it may perform better in a thread pool than other Python request libraries.
+**Using the synchronous API**. We haven't benchmarked the synchronous API. However, we do release the Python [Global Interpreter Lock (GIL)](https://en.wikipedia.org/wiki/Global_interpreter_lock) for all synchronous operations, so it may perform better in a thread pool than other Python request libraries.
 
 ## Equal performance