You are working on a text summarization project and have tested several models. Below are the ROUGE-1, ROUGE-2, and ROUGE-L scores for different models:

Model	ROUGE-1	ROUGE-2	ROUGE-L
Model A	0.55	0.43	0.48
Model B	0.60	0.45	0.52
Model C	0.62	0.46	0.55
Model D	0.58	0.44	0.50

Given that ROUGE-1 measures unigram overlap, ROUGE-2 measures bigram overlap, and ROUGE-L focuses on the longest common subsequence (LCS), which model should you select for this summarization task if your goal is to prioritize overall summary quality and coherence?