Re: Huh?
As there is now a lot more to lose now...
Oracle has has nearly all their patents invalidated on appeal, the 104 has been invalidated but is not yet 'final'. The final patent the '520 is not expected to apply and runs out soon.
The final claims are of copyright over the APIs (and the 11 unused files). Now this has a significant impact to the whole of the software industry, if Oracle were to win a claim in this area.
Google would not want to settle out of court unless there was a full admission that there was no infringement and the claims withdrawn with prejudice (difficult to make a settlement then). Oracle don't want to delay as there is a chance that there will be no patents left for them and therefore their purchase of Sun is written off to an extent.
An already shaky Oracle (in the eyes of some of its shareholders) would have to accept that the lawsuit was bogus, it has cost it millions, lost its 'prime' assets that they bought from Sun in the meantime and had to take regularly their eye off their core business for a couple of years. In this case they have put all their eggs in one basket because they have claimed that the six patents are the core patents that are worth nearly all the cost of the patent portfolio from Sun (so as to astronomically raise the potential damages of their claim). As these 6 are looking to be worth very little if anything then the whole portfolio is also worth very little.
You're right that the money is now (even admitted by Oracle) very little in the scheme of things Oracle's only hope is that they win the case on every count - they could then look at a claim of less than $50 million. This is by winning with a patent that has been invalidated (but not 'final') and claiming copyright on their APIs - unlikely. Even Oracle are unlikely to expect to win this.
This case is very similar to the SCO saga that wiped out the whole of SCO - and they do have something in common. Their lawyers are from that same 'maverick' firm - Boies Schiller.