Unfortunately in xhigh reasoning effort it will burn through 2K tokens before it has even finished its bullet point overview. It really is intense and obsessive. You might need ten times more!
Your strategy would likely help in medium reasoning effort (because there it gets caught up in the very typical Qwen looping).
Not seen looping in the “low” reasoning effort mode.