WSS alerts  aren’t on the same schedule as SPS alerts. They are more or less fired immediately

Architecturally SPS alerts have to be the result  of a crawl. The “notifier” (which is what I’ll call it because I can’t remember what the PAM acronym stands for right now.) takes it’s input from the Indexing/Gathering process.

So … what are the issues that you’re seeing?

We were having a problem with losing half of our alerts  and wondering
where to start. One guy wanted to adjust the crawl  frequency and
another wanted to use stsadm  for a WSS site as we initially didn't
understand the difference in how the alerts were handled. But based
on other research we did, you are are exactly right.

FYI we solved the losing alerts issue by turning on SMTP on the
Sharepoint server instead of just pointing to another dedicated SMTP
server to gain the benefit of the out queue. If the connection
wasn't available at the moment the alert  was generated, the alert was
lost. With its own SMTP service, the alert email would be held until
the connection was available.

I generally put SMTP on the SharePoint farm, lock it down to my servers and then forward to the corporate SMTP server. It allows me to do things like eat a set  of alerts  if I know that people are going to get alerts they shouldn’t (if an index needs to be reset or similar.)