inputs.linux_cpu fails on startup inside a virtual machine

Nobody has claimed this yet.

### Relevant telegraf.conf

```toml
# Provides Linux CPU metrics
# This plugin ONLY supports Linux
[[inputs.linux_cpu]]
## Path for sysfs filesystem.
## See https://www.kernel.org/doc/Documentation/filesystems/sysfs.txt
## Defaults:
# host_sys = "/sys"

## CPU metrics collected by the plugin.
## Supported options:
## "cpufreq", "thermal"
## Defaults:
# metrics = ["cpufreq"]
```

### Logs from Telegraf

```text
2025-06-05T08:50:29Z D! [agent] Initializing plugins
2025-06-05T08:50:29Z W! [inputs.linux_cpu] Failed to load property /sys/devices/system/cpu/cpu0/cpufreq/scaling_cur_freq: file with CPU property does not exist: "/sys/devices/system/cpu/cpu0/cpufreq/scaling_cur_freq"
2025-06-05T08:50:29Z W! [inputs.linux_cpu] No properties enabled/loaded for CPU 0
2025-06-05T08:50:29Z W! [inputs.linux_cpu] Failed to load property /sys/devices/system/cpu/cpu1/cpufreq/scaling_cur_freq: file with CPU property does not exist: "/sys/devices/system/cpu/cpu1/cpufreq/scaling_cur_freq"
2025-06-05T08:50:29Z W! [inputs.linux_cpu] No properties enabled/loaded for CPU 1
2025-06-05T08:50:29Z W! [inputs.linux_cpu] Failed to load property /sys/devices/system/cpu/cpu2/cpufreq/scaling_cur_freq: file with CPU property does not exist: "/sys/devices/system/cpu/cpu2/cpufreq/scaling_cur_freq"
2025-06-05T08:50:29Z W! [inputs.linux_cpu] No properties enabled/loaded for CPU 2
2025-06-05T08:50:29Z W! [inputs.linux_cpu] Failed to load property /sys/devices/system/cpu/cpu3/cpufreq/scaling_cur_freq: file with CPU property does not exist: "/sys/devices/system/cpu/cpu3/cpufreq/scaling_cur_freq"
2025-06-05T08:50:29Z W! [inputs.linux_cpu] No properties enabled/loaded for CPU 3
2025-06-05T08:50:29Z E! [telegraf] Error running agent: could not initialize input inputs.linux_cpu: no CPUs detected to track
```

### System info

Ubuntu 22.04, telegraf 1.34.4, inside qemu

### Docker

_No response_

### Steps to reproduce

1. Add inputs.linux_cpu to a telegraf config
2. Run inside a virtual machine
3. Error
...

### Expected behavior

Expected behavior: Even if it can't monitor all CPUs, it won't stop the initialization of telegraf. I'd expect an error to be logged at most. A bit related to this PR: https://github.com/influxdata/telegraf/pull/11313. If there are sensors or kernel modules missing, I still want monitoring. Our hardware and software is diverse, I can't generate configurations for all of them.

### Actual behavior

Telegraf initialisation is stopped

### Additional info

_No response_

Start at the inputs.linux_cpu initialization path and reproduce the failure in an Ubuntu 22.04 virtual machine under qemu. Inspect how missing CPU properties are handled when no CPUs can be tracked, then run the plugin's existing tests or a VM reproduction. Done means missing properties do not stop Telegraf initialization and the condition is logged appropriately.

Written by the indexing model from the issue text.