I think it is hard to argue that group selection has been anywhere near as productive as kin selection. However, group selection enthusiasts have been massively outnumbered over most of the recent history of the two ideas. Group selection defenders can claim that group selection has been less productive due to fewer man-hours expended on it.
The primary strengths of kin selection as I see them:
- Kin selection makes extensive use of relatedness (a concept group selection enthusiasts rarely mention). This facilitates making quantitative predictions.
- Kin selection has "Kin" in its title, which emphasizes the significance of kinship to the effect. Calling the effect "group selection" fails to to emphasize kinship. In practice this leads to applications where kinship isn't involved - and such applications are more often than not errors and mistakes.
- The field is more mature, established and respectable.