[DO NOT FIX OR CHANGE] docs: pipeline: inputs: podman-metrics: document disk I/O counters in supported-counters list

**THIS ISSUE IS FOR CLOUD NATIVE PDX EVENT, SHARING FIXING LIVE AT EVENT, PLEASE DON'T CLOSE OR FIX!**

## Summary

fluent-bit commit 80139e1ef ("in_podman_metrics: add per-container disk I/O metrics") added four new Prometheus counters to the podman_metrics input, read from the cgroups v2 `io.stat` file and summed across block devices:

- `container_disk_read_bytes_total` — Container block I/O bytes read
- `container_disk_write_bytes_total` — Container block I/O bytes written
- `container_disk_reads_total` — Container block I/O reads completed
- `container_disk_writes_total` — Container block I/O writes completed

On cgroups v1 hosts these are reported as invalid and skipped.

PR #2608 already added these four metrics to the sample Prometheus scrape output in `pipeline/inputs/podman-metrics.md`, but missed the separate "Exposed metrics" summary section further down the same file, which still only lists the memory/CPU/network counters. The two sections are now inconsistent within the same doc page.

## Where to fix

In `pipeline/inputs/podman-metrics.md`, under `### Exposed metrics` → "Currently supported counters are:", add four bullet items after `container_network_transmit_errors_total` and before the closing "This plugin mimics the naming convention of Docker metrics exposed by `cadvisor`" line:

- container_disk_read_bytes_total
- container_disk_write_bytes_total
- container_disk_reads_total
- container_disk_writes_total

Single file, single plugin, 4-line addition.

## Source reference

- Commit: `80139e1ef` in `fluentbit/fluent-bit`, `plugins/in_podman_metrics/`
- Prior partial fix: docs PR #2608 (added the metrics to the example output block only)