Welcome, yoji!
There are many possible reasons why a student's test scores might be uneven across domains. None of her scores are in any way below average, so this is really a case of relative strengths, rather than weaknesses. Her data are pretty consistently strong in math and nonverbal areas, and average in language-based reasoning and reading. Among the possible explanations might be:
1. English not first language/dual language learner. If she is hearing and reading English only in school, or if she is simultaneously acquiring two languages, it will take her a little longer for testing to accurately reflect her verbal thinking skills, and for reading achievement to catch up to her potential. This is not a problem--more of an artifact of the growth curves for dual language learners being a different shape from those of English monolingual learners. I wouldn't actually do anything about this. If there's a second language at home, continue using it naturally, let school and the community take care of English, and give her the benefits of a rich language environment in more than one language. (The longterm data on multilingual learners finds that those who have strong language skills in two or more languages have better higher level language skills than those with only one, equally strong, language.)
2. Asynchronous development. All children learn on their own trajectory, with varying rates across different domains at different times. Developmental expectations are based on population averages, on the whole, and may not describe a particular young child, even without any learning challenges. Just as toddlers sometimes go on food jags (where they will, say, eat only peas for two days, and then only chicken for a day, and then only pasta for three days), young learners sometimes have learning jags, where they are just more drawn to learning certain types of skills for a while, before they move onto something completely different. It may be that in six months, she will suddenly dive into reading voraciously, and this profile of achievement will change. There is nothing wrong or abnormal about that; it just doesn't fit neatly into our institutional schooling systems.
3. Underlying learning preference. It could indeed be that her intrinsic learning strengths slant toward quantitative and nonverbal/visual skills, and not as much toward verbal skills. Whether this constitutes a need requiring intervention depends mainly on whether it distresses her, and whether it impededs her access to other valuable resources, experiences and learning. If it does not, given that it is at least on-to-slightly-above grade level, I would choose to conceptualize it as more of an exceptional strength in math than as a weakness in reading.
And finally, I am a whole-hearted proponent of no homework for young children, and only sparing, thoughtfully constructed and assigned homework for older children and adolescents.
