Amazing! Seems like the much smaller size gives you less surface area pressing against your ear canal, less friction against removal. You don’t have the pull string/rod that would position itself in a way to block removal, as I hypothesize for my mold and I don’t see a very much of a notch/protrusion that could catch the crux helicis, as I also hypothesize for my mold. (update:) --> Also, if one looks over the surface of my mold, it’s "undulating (see along bottom edge, in particular, i.e., shallow hills and valleys that when inserted would match and grip reciprocal undulations in the increasing and decreasing width of my ear canal along its path. Your mold seems overall much more smoothly regular along its surface, which goes along with the suggestion of others in this thread that you might need an external retaining device like a canal lock integrated into the mold if there were some reason making the mold go deeper into the canal, or any of the other possible features of my mold, would not help much. (end edit).
On p.132 of his 2012 HA book, Dillon states, “No matter what style of earmold or earshell is selected, there must be a retention region somewhere on it.” He goes on say, too small a retention region or one not sufficiently angled against the exit motion, earmold slips out. Too large a retention region, difficult to insert earmold. So there is a “Goldilocks effect” in designing an earmold. Best retention design, not too big, not too small, just right based on your ear lobe/ear canal topography.