The empirical literature on the relationship between growth and inequality is a mess, because of the obvious endogeneity problem, the measurement issues and the fact that cross-country regressions are always reduced form, at least as far as I know. And the literature is inconclusive on the causation and its sign. Maybe a bit of theory would help here.
Gustavo Marrero and Juan Rodríguez argue that the observed inequality is the result of two inequalities: inequality of opportunity and inequality of returns to effort. The first has a negative impact on growth, because good entrepreneurship opportunities get lost. The second has a positive impact on growth, because what really matters for growth is the success of the most productive people, and this encourages investment in physical and human capital. Marrero and Rodríguez verify these conjectures using PSID data and compare across US states, and it works. No wonder people find inconclusive results if they lump all inequalities together.