This article about the study:

Aragón conducted a study on farm productivity of more than 4,000 farming households in Uganda over a five-year period. The study considered farm productivity based on land, labour and tools as well as yields per unit area of cultivated land. His findings suggested that even though yields were higher for smaller farms, farm productivity was actually higher for larger farms. Similar research in Peru, Tanzania and Bangladesh supported these findings.

And then the Actual Study HERE:

What explains these divergent findings? Answering this question is important given its consequential policy implications. If small farms are indeed more productive, then policies that encourage small landholdings (such as land redistribution) could increase aggregate productivity (see the discussion in Collier and Dercon, 2014).

We argue that these divergent results reflect the limitation of using yields as a measure of productivity. Our contribution is to show that, in many empirical applications, yields are not informative of the size-productivity relationship, and can lead to qualitatively different insights. Our findings cast doubts on the interpretation of the inverse yield-size relationship as evidence that small farms are more productive, and stress the need to revisit the existing empirical evidence.

Meaning the author is advocating for more scrutiny against the claim and against land redistribution as a policy stance with the intention of increasing productivity.

First, farmers have small scale operations (the average cultivated area is 2.3 hectares).

The definition of "small family farms" in this case is on average more than 5 acres, which would absolutely be under the umbrella of subsidized industrial agriculture in developed nations.