Filling missing values, two different approaches

Hello, I think it is a simple question, I was solving the chanllenge at the end of this section.

https://app.dataquest.io/m/370/working-with-missing-data/8/filling-unknown-values-with-a-placeholder

And, I wondered. Is there any difference between the two codes below?

For this particular case, and based on running both the codes, there seems to be no difference.

It is possible that, depending on the underlying code, using mask() might be a faster and more efficient approach for larger dataframes. Or it might allow for more freedom in terms of what kind of changes can be applied instead of a simple string.