In de Sitter space with a positive cosmological constant $\Lambda$, it is usually said that gravity is smaller than the value given by $1/r^2$, because $\Lambda$ *reduces* the attraction of masses at large distances.
However, many people (Verlinde, Milgrom, etc.) argue that a positive $\Lambda$ *increases* gravitational attraction at large distances. For example, MOND (with its new constant $a_0$) conjectures a higher gravitational attraction than $1/r^2$ (at large distances).
How can this apparent contrast be resolved?