This implementation is using "a standard (3 piece) road triple" instead of a more typical set of tandem cranks. The advantage of using a standard triple and mounting the Gates ring on the outside is a lower Q-Factor (-7mm in the case of FSA SL-K cranks) at 157mm, plus the elimination of 1 spider resulting in some weight savings, plus ability to use a proper granny gear (anything mountable on a 74BCD).

With a standard (3 piece) road triple, there is insufficient clearance between the granny position and BB shell. The Gates belt takes a lot of room to achieve the required clearances. The only way to achieve that is to use a set of cranks with an even longer axle and thus a wider Q-factor, and hope the ring w/belt clears your chainstays (enter Paketa's custom skinny yoke).

Additionally, to mount a Gates ring on the inside of a standard triple, you must 1) either use a special 74bcd-130bcd adapter, or mount 3 rings to a single 130bcd spider, which means that 2) you eliminate the possibility of using a drive ring of less than 38T, ie: no granny!

Regardless, discussing the merits of inner vs outer belts is pretty much out of scope here. The question was regarding potential technical issues with the given approach as shown.

One addnl thought I had about this setup... the actual Gates ring distance from the tandem's centerline is not much different than a typical left-side-drive. Surprising, but it seems true. For example the last set of tandem cranks we had, used a 49.9mm left-side timing ring chainline, whereas this SSD standard triple has 45mm centerline + outer position + 3mm spacers. If one claimed benefit of SSD is a reduction of frame and BB twist, then perhaps a few mm increased centerline is maybe not a big deal. Comments?