(preprint) Comparative Evaluation of Multimodal Large Language Models for Technical Content Simplification and Visual Interpretation