I read portions of the Court decision, and while the context is different, I think some of it could be seen as a shot across Google's bow in terms of their cache copy situation.
To put Google in a context that is more directly applicable to this decision, let's rearrange the facts with respect to Google slightly, in order to be sure that we are comparing apples to apples. It will help focus our thinking.
Let's assume that you as a webmaster hold the copyright to your website, and that includes html files, dynamic files, and images. Then one day someone named Google, Inc. comes along and asks for permission to put your entire site on www.google.com, because they are featuring certain cool sites in the category that you specialize in. They want to copy everything, and feed it right from Google. They don't want to merely link to you.
You sign a contract with Google and give them permission to do this. So far, so good. You are getting so much fame and fortune out of this that you don't even mind the fact that Google wrote exclusivity into the contract, so you are happy to take your own site offline and run off to start a different site.
Then Google starts taking the individual pieces of your site out of context, and makes them available as isolated entities through their search engine. They don't ask for your permission to use your material in this new context. Your discrete files and discrete images are now offered by Google through their search results.
It would seem that the Court decision would disallow Google the privilege of doing this.
Now here's the real situation: Google never even asked you for permission to take your whole site, but instead moved directly to the new context. Google will argue that you have an opt out available through the robots.txt and the META tag. But I think the Court might hold that in the case of a copyrighted website, the opt-out is insufficient, and an opt-in is required. In other words, the robots.txt works backwards, and the lack of a robots.txt exclusion in your root directory is not the same thing as you granting Google the right to copy by signing a contract with them.
The manner in which the cache copy is used by Google is a further disruption of your rights, because it fragments your work as a website author and violates your copyright.
So yes, I think Google is walking on thin ice with their cache. I suspect the Court would allow grabbing the files for indexing purposes, but not making a separate, complete cache copy available on their site. It might be different if a nonprofit was doing it, as the "fair use" provisions do allow some slack on the grounds of a) the amount of material copied, and b) whether it is for purposes of profit. For example, libraries are given a fair amount of slack, which is why you can find copying machines in libraries.
Google loses on both counts. Their database is massive, and they are trying to make a profit.