LookupQueue can be refactored into a memory-managed queue

### Search before asking

- [x] I searched in the [issues](https://github.com/apache/fluss/issues) and found nothing similar.

### Motivation

`LookupQueue` could be refactored into a memory-managed queue similar to `RecordAccumulator`, which would significantly improve the efficiency of lookup batching.

### Solution

_No response_

### Anything else?

_No response_

### Willingness to contribute

- [ ] I'm willing to submit a PR!

Start by locating the LookupQueue and RecordAccumulator implementations and the lookup-batching entry points that use them. Compare how each manages memory and batching, then define the refactor's completion criteria around equivalent lookup behavior with the intended memory-managed batching efficiency.

Written by the indexing model from the issue text.