neuvector: see if sylvactl unit timeout needs to be refined for update scenarios

The following discussion from !8108 (merged) should be addressed:

• @tmmorin started a discussion: (+2 comments)

  For DaemonSets there is a known pattern : they start in parallel, but typically, with default settings they are upgraded sequentially, which takes a time linear with the amount of nodes.

  What we can do in such a case is compute sylvactl timeout with a formula multiplying the time to upgrade a pod with the number of pods (_internal.node_count if the DaemonSets is on all nodes).

  (neuvector has daemonsets, right?)