DOC clustering speed with connectivity matrices (scikit-learn#27020)

Author: rprkh

examples/cluster/plot_agglomerative_clustering.py

@@ -7,7 +7,7 @@
 neighbors.
 
 There are two advantages of imposing a connectivity. First, clustering
-without a connectivity matrix is much faster.
+with sparse connectivity matrices is faster in general.
 
 Second, when using a connectivity matrix, single, average and complete
 linkage are unstable and tend to create a few clusters that grow very