Short answer, useful, but not that much.
Long answer, from my particular case, although it won't help you much:
We have about 180 servers, 38 of those are Unix. Three older AIX boxes, running a legacy in-house application that refuses to die, and 35 Linux boxes, mainly proxy servers, and a handfull of application servers in the middle. The application servers are virtualized, the proxies are all physical servers.
We don't even have the AIX boxes in Lansweeper, so no help there, and for the Linux boxes, Lansweeper is used just for inventory. We get overall info like brand, model, warranties, that we can use for inventory purposes, but not that much else.
Outside of Lansweeper scope, all of our servers are running CentOS, I'm now upgrading the CentOS 5 machines to CentOS 7, pending the EOL on March. For all the servers, we use a kickstart minimum build, and we keep the servers updated using yum. The application servers are a bit more tricky, but since they are just a few, not that much work... We do patches and vuln scanners quartely, so they are always more or less updated.
We currently don't use any kind of patching/management tool (cfengine, puppet, chef), but our growing number of machines has us thinking of doing something... A former unix admin knew Puppet and did a little demo, but he left the company and no one picked that up... I would say we just crossed that line where the ROI of learning and deploying a tool like that is head to head in time of doing thinks manually...
Finally, even on Windows, we don't use Lansweeper to manage software deployments, just for one-off tasks.
We use a mix of WSUS/Retina for Updates and Patch Management, and EMCO for in-house software and applications not know to Retina, those are tools we already owned and used before LS had the deployment module...