According to the spec, "when dfdl:lengthKind is 'explicit' (and dfdl:length is an expression), ... the data value is padded to the length given by the XSD minLength facet"

I wrote a test to verify this, and the unparser is padding the data to the length expression, not the minLength facet.

See test
unparseDelimitedPaddedString10
in files:
daffodil-test/src/test/resources/edu/illinois/ncsa/daffodil/section13/text_number_props/TextNumberPropsUnparse.tdml
daffodil-test/src/test/scala-debug/edu/illinois/ncsa/daffodil/section13/text_number_props/TestTextNumberPropsUnparseDebug.scala
daffodil-test/src/test/scala/edu/illinois/ncsa/daffodil/section13/text_number_props/TestTextNumberPropsUnparse.scala