The conclusions drawn in Inoue & Matsuzawa's (2007) study, which seems to be available here, are suspect. First off, the sample sizes (6 chimp, 9 human subjects) are simply too small to draw good inferences about working memory, at least about human working memory, but as I imagine chimps are somewhat expensive subjects, c'est la vie.
Secondly, and more worryingly, the study design itself seems to have a real problem: Matsuzawa does not reveal how many hours of practice either group of subjects have had, but presumably, the humans have not had any practice, while it is explicitly noted that the chimps have had some. The mother-offspring pairs practiced the numerals task from offspring age 4, and then the block task from offspring age 5. Even if we ignore the sample size problems and take the reported results at face value, we are still left with a situation where the mother chimp Ai underperformed the humans, who in turn underperformed the offspring chimp Ayumu--also suggestive of a training effect.
This strikes me as a problem for Matsuzawa's claim, because we know that the ultimate limit of task-specific human memory span is very high compared to a task-general baseline. For instance, Ericsson, Chase and Faloon (1980) reported a case study of a subject who increased their memory span from 7 to 79 in 230 hours of practice, and one need look no further than the world memory records to see that the lack of control for practice seriously undermines Matsuzawa's claim about this unexpected "relationship" between chimp and human WMC.
A more restrained interpretation of the evidence presented in the article would suggest that sufficiently well-trained chimps can outperform untrained humans on the particular task. Any conjecturing about "evolutionary trade-offs" on the basis of these findings seem unwarranted to me, but that would have been the case even if these results had been ironclad.
References
- Ericsson, K. A., Chase, W. G., & Faloon, S. (1980). Acquisition of a memory skill. Science, 208(4448), 1181–1182. doi:10.1126/science.7375930
- Inoue, S., & Matsuzawa, T. (2007). Working memory of numerals in chimpanzees. Current Biology, 17(23), R1004-R1005.