"Java’s not worth building in [to the phone]" and
"It’s this big heavyweight ball and chain."
Many people have taken this to be an attack on Java ME. However, I see it a bit differently. You have to put this into context.
The iPhone is playing catch up. The iPhone hardware and software teams have had (and still have) a lot of technologies to put together. They obviously evaluated Java ME at some stage and correctly identified that it would take a very large amount of effort to support the current (and future) JSRs. Over a relatively longer period of time, OS and phone OEM companies in the Java ME ecosystem have expended (and continue to invest) very significant amounts of effort to support Java ME. I believe it just wasn’t seen to be worth (financially sensible) porting onto the iPhone and would have represented an onward financial ‘ball and chain’. Hence, Apple may have arrived too late to join the Java ME club – because late membership has become too expensive.