After reading his books Tartessian (2009, 2013) and Tartessian 2 (2011), I've come to the conclusion Koch's work is an unmitigated disaster. One gets the impression he tries to get a Celtic interpretation of Tartessian segments at all costs, no matter the inconsistences in which he might incurr.
Despite so, some lexical isoglosses with Gallaecian (Western Hispano-Celtic), Lusitanian as well as Iberian are apparent:
aibu- in the anthoponym Aibu-ris ~ Iberian aibe-. Possibly the Lusitanian counterpart of Celtic *aidu- 'fire'. Less likely a relationship with Celtic *albijo- 'upper world' (cfr. Gaulish Albio-rix).
Haitᵘura /waidura/ ~ Iberian baiduŕa (the transcription of rhotic signs is reversed in the NW Iberian with regard to the S and SW scripts). Probably related to Iberianbaides 'witness' < IE *weid- 'to know, to see (as a fact)' (cfr. Celtiberian ueizos).
niirabºo /nērabo/ ~ Basque nerabe 'servant; bachelor; child, boy'. From IE *nēr-'(young) man' (cfr. Celtic *nero- 'hero').