Fix temperature support check

[eosti]

An odd issue I noticed on my EdgeRouter X (EdgeOSv2.0.8-hotfix.1): it doesn't support temperature checking, so when I enabled the default config, I assumed that the script would automatically detect that and give a nice error message, like the power monitoring does. Instead, it threw some weird error about a metric parse error. I found two issues:

1. The error message for the temperature is Temperature not supported on this platform. Compare that to the one for power: Power status is not supported on this platform. Note the lack of "is", which would seem to cause the grep call to not function properly for the temperature check. This would be an easy and straightforward fix, except...

2. (this took a while to find) Unlike getPowerStatus, getTemp will print the error message itself, and then return null instead of returning the error message. So checking for a substring never would have worked, because temperature_info is empty when there is an error! Gotta love inconsistent code practices.

Anyway, this PR changes the check condition to check for null instead of checking for a substring. It also adds the error message that gets lost due to the getTemp shenanigans, so that the error will show up in Telegraf logs.

Bonus fix: in the default config, the firmware check uses the interval 1d, which isn't valid. I changed it to 24h which is valid and equivalent.

[x70b1]

Thanks for your work on this! Just a smaller change is required.

[eosti]

A small change: now the getPowerStatus command will log an error explicitly instead of what it was doing previously with the implicit logging via the output of grep. There is no change in how the script behaves: this is just to make it more obvious where the error might be coming from and to match style with my modified getTemp command.

[x70b1]

Very cool @eosti !