A System-Independent Digital Root Fingerprint in the Quran: Evidence from Fifteen Computational Experiments across Two Languages

v3 — adds Experiments 14-15. Exp 14 (Random Encoding Stress Test): 33 random encodings tested. 0/33 beat Special-6 (p < 0.007). Mean {3,6,9} = 31.3% (below chance). The fingerprint is NOT an artifact of arbitrary encoding. Exp 15 (Frequency Order Encoding — System 3): Third independent encoding based on inverse frequency weighting. Surah-level {3,6,9} = 21.9% (not significant). But transformation map reveals 5 self-preserving roots {1,2,6,7,9} (p = 0.0015). Each encoding system reveals a different structural layer of the text. Unified preprint: 15 experiments, 3 encoding systems, 5 control texts, 2 languages. The Quran carries a multi-layered digital root structure absent in all controls.