Good luck with that here.
There's a huge backlog of fair-use cases for them to base their decision on. I don't think it would be particularly difficult or nasty in this case, given that the ability to copyright APIs was a presumed fact during the trial. Also, the judges do not all need to agree; majority vote wins. So the moment they feel that further debate is unproductive, it's time to vote and move on.
It is going to have be be very clear and something that can be ruled on in a short space of time. There is no way you can argue at all that anything clear has been missed here, or argue in clear terms how things have been weighted.
Stuff like arguing that commercial reasons haven't been weighted properly is entirely subjective, because it depends on arguing how transformative use is - and that is what has happened here. The jury have already decided on that one. The 'backlog' of fair use cases will simply throw up the same subjective issues.
An application like that to a court is not going to impress any judge one iota. You can't just wander back into court and argue "The jury has been unreasonable" without some totally solid evidence. The jury also ruled unanimously, so it wasn't even close.
I can certainly see an argument here that certain facts, such as commercial use, were not properly considered by the jury. Commercial use is typically a very large factor in fair use, and should receive a correspondingly heavy weighting.
Also, remember that this would merely permit a retrial. It would not be an immediate ruling for Oracle, as one normally thinks of appeal decisions. It would be a big reset button on the whole process.