Okay, though it appears not, the entry for *tegos you post actually proves my point; the stem is *tegos-, you notice, not *tego- as it would be if it were an o-stem and not an s-stem... I don't understand why apparently the same publication (what is it?) then quotes *tego-slowgo-. Raimund Karl in Thoughts on the evolution of Celtic societies and grand Celtic narratives has *tege- as the compositional form and I've assumed he was correct. The word for horse *marko- is an o-stem, but we're talking riders, not horses. The Welsh plural marchogion suggests Brit. *marcacones rather than *marcacoi; this is my presumption... I don't know why the authority you quote goes for an o-stem agentive *-ako- rather than *-aku- when the Welsh suggests otherwise. Maybe there's something about Old Welsh plurals (and it's still -ion/-yon in Old Welsh) that I don't know about and which causes o-stems to have n-stem plural endings put on?