ken_b, thanks for the title suggestion. I've changed the wording to include date and year.
Also, worth noting in new serps which seem to parallel the release of these updates... No direct link to specifics in the guidelines, but nevertheless, I've felt that there's been a shift in the serps, in very broad terms, as the guidelines suggest that's where Google might be going.
Eg... the guidelines are getting much more granular, and I feel I'm seeing that in the serps as well. I'm noticing results that are increasingly granular and nuanced withn a document, and competitively within a niche. Google seems to be digging deeper. This is a combination of observation and speculation, based on personal searching....
I should note that these are mostly research searches, not highly competitive results for large mass markets, but I'm nevertheless discovering, eg, established brands of "widgets" that I previously hadn't known existed... These brands are not on pages as featured items with dedicated titles, but simply as one of say 4 or 5 professionally accepted brands that were being discussed... often in forum discussions.
A widget might be (hypothetically) described, say, as something like: ""Whatiso', a leading Japanese brand, never widely advertised, which found its market among professional users who felt that the leading Xyz unit made too many demainds on workflow, but who nevertheless wanted comparable results at a lower price without too much compromise...." etc, but that and the context seems to be enough for Google to pick up on it. I'm also getting amazed at Google's ability to dig into the meaning of a page by the use of context.
This is vivid contrast to some (not all) results in the past 10 months or so which have frankly been awful. Often Google would have a choice of say two different words in the same title or article and needed to make a choice of which of these two was the tail and which was the dog... ie, which was the main subject. A good percentage of the time, Google was getting it wrong. What struck me as odd about this is that Google had the same distinction nailed down correctly for years... which has made me think that Google has made some basic changes.
I've recently been seeing this situation improve considerably as Google has been able to train on more data... but it is now AI/ ML - driven. In some cases, results have evolved as if Google were learning the basics all over again, and in many situations, these were choices which I'm certain Googe had sorted through before. Eg, I've seen discussion of the return of domain clustering... ie, multiple results together from the same site, as Google winnows these down to one or two, again something which Google has done before. This makes me think that Google might be using some basically new algorithmic strategies which need this kind of re-testing... and once again must check it's choices not only among sites, but among mutltiple pages on the same site where a single query dominance is strong enough to deserve evaluation.
This idea of a new algo strategy is speculation, but the serps feel like we're seeing the return of certain types of results that Google had previously used for its own style of multi-variate testing. It feels now as though Google is able to test more variables at once than it had before... making for some curious serp displays... as Google hones down what to show in what situations. </end speculation>
