[Monitoring] Wrong defaults for alerts

[chrisronline]

• Disk usage alert needs to default to 80% (current code is 90%)
• Missing monitoring data alert needs to default to looking back 15 minutes (current is 5 minutes)
• Missing monitoring data alert needs to default to a throttle period of 6h (current is 1 day)

[elasticmachine]

Pinging @elastic/stack-monitoring (Team:Monitoring)