Then we can agree to disagree. There’s certainly a part of subjectivity in what one considers “complex”. I remain convinced that some algorithms are better expressed with mutation, for both humans and compilers.
It is unlikely that there exist a single programming paradigm categorically better than all others. Following this premise, the interest of MVS (i.e., the trend mentioned by OP) is to offer a framework in which functional and imperative programming can mingle with relative ease.