There is also some psychological research by Nicola Pitchford on an intervention in Malawi where students received tablet-based instruction in math ( https://doi.org/10.3389/fpsyg.2015.00485 ). The results of kids receiving the intervention are compared to a control group receiving standard instruction and a placebo group receiving non-math tablet instruction. The sample is relatively small, since the experiment took place in a single school. There are 183 kids across three grades with one treatment arm in the first and two treatment arms in the second and third grade, plus a control group in each grade, resulting in an average 38 students per treatment arm / control group.
With that caveat in mind, Pitchford finds that "the maths tablet intervention was significantly more effective at improving mathematical attainment than current instructional practice for primary school children in Malawi. [...] Results showed no advantage for the non-maths tablet intervention placebo group over the normal practice control group [...]. Furthermore, the maths [curriculum knowledge] acquired by children in [grade three] through interaction with the maths tablet intervention generalized to novel items, that were not practiced during the maths tablet intervention, and were delivered in a different, paper and pencil, format."
I'm not an expert on the methodology used (since it not standard econ stuff), and since effect sizes aren't presented in standard deviations, I find it hard to compare this to results I know from economics. (Although Pitchford notes that the "large effect sizes found in this study for the children exposed to the maths tablet intervention exceed, or are comparable to, those reported by other studies in the literature (Räsänen et al., 2009; Shin et al., 2012; Praet and Desoete, 2014) conducted with European and North American primary school children.") The program is being scaled up, so more research may be yet to come, looking at more students and more schools. But from this initial evaluation, it seems this might also turn out to be a "computer-assisted learning program that delivers learning results", and could be worth keeping an eye on.