fix(mastodon): set scrape interval explicitly to null

This patch drops the custom scrape interval and replaces it with the
default from the cluster's Prometheus instance. This helps to keep
metrics better in check and resolves the issues with trying to calculate
rates over 1m in the dashboard.