I've tried to run the bot with the shaman sample at higher rank (7).
I used a full sea giant deck. The bot made only missplays, and AI was horrible.
At low ranks, for golden portraits I think the bot is ok, while still doing some major miss-plays which make it a bit slow. But at this stage, getting past rank 10 without spending hours configuring an excel file to precise each and every behavior, I think the bot is far to be viable.