antigravity interactions harness - improve TTFT (per-turn -> per chunk streaming)

## Problem statement

The Antigravity **Interactions** harness's output granularity is **per turn**, coarser than the SDK harness (which is per-Content / block).

## Code pointers

- Buffer text deltas into one string: https://github.com/google/ax/blob/6c6bf181062d3e336fa209f4ac4dc6258d02c219/internal/harness/antigravityinteractions/antigravityinteractions.go#L921-L933
- Emit once per turn (`emitText` at the top of the Run loop): https://github.com/google/ax/blob/6c6bf181062d3e336fa209f4ac4dc6258d02c219/internal/harness/antigravityinteractions/antigravityinteractions.go#L377-L381
- `emitText` builds a single `TextContent` per turn: https://github.com/google/ax/blob/6c6bf181062d3e336fa209f4ac4dc6258d02c219/internal/harness/antigravityinteractions/antigravityinteractions.go#L422-L433

## Desired behavior

- Emit the first visible output as soon as the SSE stream produces it, rather than buffering the whole turn.
- Stream per chunk / per Content instead of per turn to improve TTFT.
- Keep final/durable content semantics correct.