Update on web host situation: I signed up for a new hosting account at a to-be-named host after a great introductory chat with them, but the order became effed up due to the fact that I was making the order from a Canadian IP address with my US address/credit card information. So I'm still working on that issue - this will be after this weekend then another few days at the cottage where there is a blissful lack of internet access.
I.P.'s points regarding database access are well-taken - except for the fact that the forum is so much less trafficked than the main blog*. After dropping the SimpleMachines posts-per-page number down, I still saw a couple of full-website drops in the following days, even while Google Analytics indicated that virtually nobody was using the forum at the moment. But these are just initial speculations - I'm still having fun learning about it.
Does anybody know the shell or cPanel method to check the CPU time that has been tied up by the database access for each of these separate parts of the site? I'm guessing there are two different measurements - CPU time used on my actual host by wordpress and simple machines script processes, plus database server time (on another machine somewhere) from the queries my main server is making. (Unless it all runs in one space, partitioned off for me?)
Anyway, the big news is that I turned off the main blog's "all posts since the beginning of time" feature and turned it into a static page instead. That was an obvious resource hog, since it is a very popular page and it does a gigantic loop through the database for every person who reads it. You may laugh at the fact that I allowed that script in the first place, but I guess I was assuming that computers were fast enough these days to not bog down doing a few hundred loops of a "for" statement, a few tens of thousands of times per day. I know, I know, it's all in the number of CPU instructions that REALLY happen for each database query (probably millions), but still.. this blog is TINY compared to a page like the Wall Street Journal or even Lifehacker.com, so I'm surprised that it is slowing down with only 60 pageviews per minute. I mean, even a large-print paperback book could deliver almost as many pageviews per minute to a group of 60 people!
Anyway, it's all learning for me, which makes it all good..
*There's also a nice "uncorrelated assets" thing going on with the forum where usage ramps up in the evening hours and on weekends, while the main blog traffic is driven purely by people slacking off at work: mostly in the US, mostly during the 9-5 period in the country's four primary time zones.