This report synthesises findings from 13 peer-reviewed papers addressing the following research question: How do multimodal models trained with synthetic image-text pairs perform on cross-domain alignment tasks compared to models trained on natural data when evaluated using CLIP scores versus downstream. 9 claims were extracted from source literature; 9 were independently verified against retrieved documents. An automated multi-reviewer quality assessment produced a score of 8.7/10. This report is a machine-generated literature synthesis and does not constitute original research.Research goal: How do multimodal models trained with synthetic image-text pairs perform on cross-domain alignment tasks compared to models trained on natural data when evaluated using CLIP scores versus downstream classification accuracy?Autonomous literature synthesis. Automated review score: 8.7/10. Full text and citation available at Assignee Research.