Aye, there have been numerous investigations into it. Among linguists, it seems to be basically taken as a given that homophones, particularly when compared to less skilled native readers, are going to be easier for the non-native, because of acquisition methods. I should've prefaced that with the disclaimer that I am not a linguist, but I have worked in the past with linguists on language-related work, both in L1 (native) and L2 (non-native) contexts. Here are a couple of references I was given on this specific subject:

1. Van Orden, Guy C. (1987). A ROWS is a ROSE: Spelling, sound and reading.

2. Binder, K. & Borecki, C. (2007). The use of phonological, orthographic, and contextual information during reading.

3. Ota, M., Hartsuiker, R. J., & Haywood, S. L. (2010). Is a FAN always FUN?

Full texts for all of those are available online.