What an SRT file can't represent
A WebVTT file can include elements an SRT has no equivalent for. Positioning settings tied to the timestamp (line, align, position) can't be translated and are removed. NOTE and STYLE blocks aren't part of the subtitle content and aren't included in the final text either.
A particular case is speaker tags. A VTT identifies who's speaking with a tag wrapping the line; an SRT has no such mechanism. Rather than discarding that information, it's converted into the name followed by a colon at the start of the line, the usual way dialogue is represented in subtitles.