The Ansible setup module will give you facts about target systems including memory usage, uptime,swap space, it would be fairly simple to extend this with a custom fact to also grab the contents of /proc/loadavg
Reading between the lines though, it sounds like you have been tasked to use ansible as an ersatz monitoring solution? Scheduling ansible runs against a number of hosts to grab runtime stats might not be the most efficient use case, you may want to investigate dedicated tooling like Sensu, Nagios, Zabbix or similar.