I recall there was some issue in the past with hash function requiring proper ordering of tags or something.

Yes, this is still the case:

func (s *Span) Hash(w io.Writer) (err error) {
	// gob is not the most efficient way, but it ensures we don't miss any fields.
	// See BenchmarkSpanHash in span_test.go
	enc := gob.NewEncoder(w)
	return enc.Encode(s)
}

You can only compare hashes if the tags are properly sorted. Surprisingly, we don't have an explicit adjuster for that, but some of that sorting happens in ip_tag.go and sort_log_fields.go adjusters. Perhaps it would make sense to pull the sorting explicitly into SortTagsAndLogs adjuster and makes sure it is called before DedupeBySpanHash adjuster.