Although peer review has been widely used for formative assessment in writing instruction, there remain concerns about whether assessors are at a sufficient writing performance level that would allow them to identify major problems in the reviewed work and provide helpful feedback to improve draft quality. Little empirical research has examined how assessor writing performance specifically influences problem identification accuracy and helpfulness of feedback, nor has it acknowledged different grain sizes of assessor performance. Assessor writing performance at different grain sizes (i.e., performance at the levels of genre, dimension of a genre, and specific problem topic) was assessed alongside problem identification accuracy and feedback helpfulness in 234 high school students who participated in an anonymous multipeer review in a secondary writing course in the United States. A correlation analysis showed that assessor performance levels on specific problem topics were meaningfully separable, thereby allowing for consideration of the effects of assessor performance at genre, dimension, and topic levels. Multiple regression results indicated that assessor writing performance was unrelated to problem identification accuracy at any grain size. Therefore, scaffolds in the reviewing process appear sufficient to support problem identification accuracy. However, assessor writing performance, particularly on specific dimensions and specific topics, consistently predicted helpfulness of feedback, even though lower performing assessors rarely produce incorrect advice. Theoretical and practical implications of the findings are discussed.