If your goal is to A/B test two Lookalike audiences, don’t launch the $800 ad sets first. Instead, set up a clean A/B test with two ad sets at $400 each, so you can directly compare performance under the same conditions.
If you've already run the $800 ad sets, pause them, duplicate for testing, reduce the budget to $400 each, and run your test from there.
Keep all other variables (creative, placement, etc.) the same to get accurate results.