One thing i have noticed is that for sites where i lost P1 rankings for the main KW that the site was optimised for , and have totally disappeared in the serps for that KW , i still rank very well for the wordpress tags - i use LOTS of WP tags with my sites and its very interesting to see them still ranking well even though the main KW is nowhere to be seen. I am still getting reasonable traffic now from those tags and getting sales as a result. I also see my competitors who were not on P1 , now ranking for their WP tags.... definitely something in this guys... OK, so why are the 'tags' (eg www.mainkeyword.com.au/tag/secondaryLSIkeyword/) still doing well and what does it 'tell us' about the penguin update.
My conclusions are
1. The 'over optimisation' thing that google mentioned is true - as an experiment i might try ranking a new site with domain name - www.secondaryLSIkeyword and then have more emphasis for the mainkeyword ie: the H1 /H2 , content and meta title WITHOUT going over the top.
As this excellent article (Solutions To Google's NEW Algorithm Update For Over Optimization and MORE!) points out "For every h1 tag you use, plan on having 5-6 h2 tags to expand on the keyword and subjects." - spot on i think.Do not over optimise for 1 KW - lots of LSI in the content needed.
2. Next, i took a look at the backlinks for the sites that are NOW (post penguin) ranking in the top spots of P1 google.I used SEO Spyglass. It follows the same pattern - they have more diverse anchor text - ok, this is not new but its probably more important now to have very diverse anchor text. What has NOT changed is that the top sites still get links from high PR 'quality' sites , generally related to the KW they are trying to rank for. Not easy to get i know (thats why im building my own blog network, to create those sites to link to my money site from).
3. Use LOTS of WP tags , LSI related - even if your site gets hit for the main KW, you will still get traffic from the tags now appearing all over the serps in good positions.
I think by taking the emotion (and disappointment etc) out of losing hard won rankings and looking at what is NOW (post pengin) needed to rank in the serps , we can 'reverse engineer' what is needed to rank post penguin google update. All is not lost.
Sorry for the ramble - would be interested to hear what others have made from this last update - mainly from an observation point of view as to why sites have high rankings post penguine