Summary

Refactor buffering strategy implementations to use the existing ring_buffer template instead of duplicating ring buffer logic.

Parent Issue

Part of #309 (Consolidate buffering layer - C++20 Module preparation)

Current State

• buffering_strategy.h imports ring_buffer.h but doesn't use it
• Each strategy class (fixed_size_strategy, time_based_strategy, etc.) implements its own vector-based buffer
• Ring buffer logic (overflow handling, circular writes) is duplicated

Problem

• ring_buffer.h provides optimized, lock-free implementation but is underutilized
• Strategy classes duplicate buffer management code
• Inconsistent buffer behavior across strategies

Proposed Solution

1. Modify strategy classes to use ring_buffer<buffered_metric> internally
2. Keep strategy-specific logic (flush triggers, priority handling) in strategy classes
3. Delegate core buffer operations to ring_buffer

Acceptance Criteria

• Strategy classes use ring_buffer for storage
• All buffering strategy tests pass
• Performance maintained or improved
• Memory footprint reduced

Files Affected

• src/utils/buffering_strategy.h
• src/utils/ring_buffer.h
• tests/test_buffering_strategies.cpp