A problem with conducting a missed approach after a baulked landing is that you're starting from a height somewhat less than what was considered in the missed approach obstacle clearance gradient.

Having watched a fair number of circling approaches in the simulator where it often starts to go wrong is a lack of wind correction when on the outbound part of the instrument approach.

This is simply because we are not used to applying corrections in this sense since the instrument approach is usually into wind. But wait a minute - WHY are we circling? Too much tailwind for landing straight in.

This coupled with the fact the airspeed for the approach is higher (given reduced config for circling) AND we have a tailwind makes everything quite different - different pitch attitude and thrust settings (to what is the "norm").

All this can combine to make the whole circling manoevre a rushed exercise. Great care needs to be applied to the outbound part of the procedure and attention to full and comprehensive briefing for what is, for many, an "unusual" approach.

As regards the thread topic then circling minima should be bugged but your ops manual should spell this out.