The frontside, with one of the five fans initially tested:
A macro view of the modified HDT-S1283 fan mount. At the right, you can see what's been cut off. When the fan is sitting on the four rubber nubs, the nubs carry the weight and the tape only keeps the fan from sliding horizontally off the nubs. Thus, there's no stress on the tape. I've done 7 fan changes so far on the original tape and it's got life left:
The stock HDT-S1283 is flatly not designed to ever change the fan. I used diagonal cutters to cut away the undesired portion of the rubber fan mount. I guess Xigmatek (or KingWin in this case) never expect you to have to remove the fan to clean the dust off the fan and fins.
This fixture is designed to allow fan noise comparisons at equal airflow when there's a known backpressure on the fans. It works; I have the data on the first 5 fans tested. I'll post that later this evening in the User Review forum.
As I predicted, I had to wait until dark to measure the sensor RPM because of the slow rotation and hence infrequent strobe flashes.