Oh, I don't think kongresejo is seriously ambiguous.

The question is whether the root 'kongres' must be of a definite grammatical class.

Actually, I am grateful for this example, Miland. It seems to me to present a nice test of the theory of native grammtical class.