Clean up error messages in log

[harp-intel]

This pull request focuses on improving error handling and logging in both the configuration and reporting commands. The main changes enhance how errors are aggregated and surfaced to users, and adjust logging severity for certain non-critical issues.

Error handling improvements in configuration:

• Aggregates errors encountered while setting configuration flags on each target in setOnTarget, so that all errors are reported together instead of failing fast. The aggregated error messages are sent back through the channelError channel. [1] [2] [3]
• Updates the main runCmd function to collect and check for errors from all targets, printing and logging any errors after all processing is complete, and ensuring the command exits with an error if any occurred. [1] [2]

Logging severity adjustments:

• Changes several log messages from Error or Info to Warn in dimm.go and telemetry_tables.go to better reflect that these situations are not always critical failures, such as unrecognized DIMM size formats or missing telemetry data on some platforms. [1] [2] [3] [4]

Robustness in telemetry reporting:

• Adds a guard clause in getMetricAverage to return an empty string if there are no fields to average, preventing potential panics or misleading results.