My recommendation is to use graceful restarts whenever and wherever possible. This is an important setting, because else, on each reconfiguration all Apache instances will briefly stop until Apache is restarted, interrupting web server operations for a short time.
The "issues" the article means is probably parallel reloads/restarts of Apache. When you reload the web server configuration while another reload process is already ongoing, Apache can fail and needs to be restarted manually. With a restart interval that is longer than the time an Apache reload/restart needs this is unlikely to happen. It still happens on some occasions, but for these rare moments, you have a script like the Plesk built-in watchdog that can restart Apache automatically, or a home-made script that checks the Apache status more frequently. The later is what we do here, every 10 seconds, but this is entirely up to you and needs some programming.