Concurrency Budget is Static & Estimated From Hardcoded Magic Constants

Nobody has claimed this yet.

## Problem

In `StreamingRawDataset`, the adaptive concurrency budget for remote downloads was previously calculated using hardcoded static assumptions:

```python
_ASSUMED_AGGREGATE_BANDWIDTH_BPS = 100 * 1024 * 1024 # ~100 MB/s
_ASSUMED_REQUEST_RATE = 6000.0 # target req/s
_ASSUMED_REQUEST_LATENCY_S = 0.040 # 40 ms RTT
_DEFAULT_MEDIAN_FILE_BYTES = 256 * 1024
```

These assumptions do not reflect the large variation in real-world cloud environments.

* **Low-Bandwidth / Shared Nodes:** Multi-tenant EKS/Kubernetes nodes may provide only ~10–20 MB/s of effective bandwidth. The static model can overestimate sustainable concurrency, leading to **HTTP 429 throttling, socket saturation, and connection timeouts**.

* **High-Bandwidth / High-Performance Storage:** Dedicated A100/H100 nodes with multi-Gbps networking or low-latency object storage can support significantly higher throughput. The conservative static cap can unnecessarily limit concurrency, resulting in **GPU underutilization and reduced data-loading throughput**.

* **No Empirical Feedback Loop:** Actual download throughput and request latency were not incorporated into subsequent concurrency decisions. As a result, the concurrency budget remained largely independent of the performance characteristics of the current network and remote storage.

### in short:

The previous concurrency model was therefore **configuration-driven rather than measurement-driven**. A more robust approach should use empirical runtime measurements—such as observed download throughput and request latency—to dynamically estimate sustainable concurrency for the current execution environment.

Start with the StreamingRawDataset concurrency-budget logic and the hardcoded bandwidth, request-rate, latency, and median-file-size constants described in the issue. Determine how observed download throughput and request latency should influence later concurrency decisions, then verify that the budget adapts across low- and high-bandwidth environments without causing throttling or underutilization.

Written by the indexing model from the issue text.