OK, so this question doesn't get asked all the time, but it does come up every so often.
So here's the official response:
Firstly, do you really need ZAP rewritten?
Secondly, even if you think you need it rewritten, most ZAP users don't really care what language it's written in, they are more concerned about its functionality.
ZAP is a complex cross platform product that contains well over 1 million lines of code.
Java is a (note, not ‘the’) good choice for complex cross platform projects, as can be seen by the number of commercial security tools written in Java. There are quite a few other languages that could also be suitable. However the majority of ZAP developers know Java better than any other language. And a small subset of potential users will object to any specific technology, be it Java, Python, Ruby, C# etc. etc.
We could rewrite ZAP in any of the other suitable languages, but it would take us a very long time, and it would almost certainly be a worse product due to our lack of experience in that language.
We would piss off most of our users, many of our existing developers and also alienate that small percentage who hate that particular technology.
So we're not going to do it.