If you can already win consistently, why do you seek new strategies? I would seek a new game (or at least new competitors) where you don't yet consistently win.

Curiosity is one excuse. Learning more about the game-space can also generate better strategies. Strategies in a game such as Go can always be improved.

The point about playing to learn is a stepping stone to the point that losing is not something to avoid at all costs. I would rather lose and learn than win and not learn. This drives me to do exactly what you suggested. Seeking new competitors or new games is roughly equivalent to playing to learn since you are putting yourself in a position where losing is more likely. In both of these cases, losing is good because you can learn from the losses. Winning is still better but winning against losers is useless. (Unless there are prizes on the line.)

More generally, I think this is an example of exploration (learning) vs. exploitation (winning) strategies. Personally, I think that people are more risk adverse than is otherwise optimal, so we should attempt more exploration than we normally do before settling into an exploitation routine.

The major point I was trying to make is that attempting to exploit and failing at it will automatically provide a source for exploration. The idea of playing to learn is exploration, but if your exploitation routine is consistently providing exploration, why bother with the exploration routine? Play to win and use the losses to learn how to win.