Literally so: 2 of the 3 changes they highlight are "movement" within the margin of error on the estimates (in both cases, only a bit over half of the MOE). So just from sampling error, there is very low confidence that a change in the direction stated in the headline actually occurred.

A responsible headline would be "More Californians moving to Florida", without mentioning the "changes" in California-to-Texas/Arizona flows, which from the data cannot reliably be said to have occurred.