Suggested solution:

(1) Significant clusters are characterized by the coordinates of the local maximum in the associated contrast images, and not the coordinates of Tmax.

(2) T-maps are used to detect significant clusters only. Within significant clusters, color-codes are attributed to the voxelwise contrast (and not the t-value)

(3) After a cluster has been considered significant, we suggest determining the outline of the displayed region with a more 'fuzzy' approach, including information from contrast images.
The following algorithm is proposed: (1) calculate the mean contrast in the original cluster (2) use this mean contrast as a 'least significant contrast' and include contiguous voxels that exceed this contrast, provided they satisfy p<0.05 uncorrected. The primary goal of this algorithm is to support the presentation of results. No further statistical assumptions are made and we do not claim that each displayed pixel is significant in terms of rigorous hypothesis testing.