Does YAML configuration has more functionality then DataStream/Table APIs?

[matanper]

I'm going over Flink CDC documentation, it seems that using the table/datastream APIs directly has less functionality then using the YAML configuration, is that correct?
For example having to manually define the columns instead of automatically resolving them.
I want to package my job into jar, and it doesn't seem possible with the YAML configuration