Would this imply something wrong, or out of spec, with the uplifted fuel?
Don't think so. The extent of the cooling is a function of time and the extra low temps throughout their time at CRZ LVL. If they did the entire trip in a real cold pool of upper air, that could really suck the BTU's outa the airframe (i.e. fuel-tanks). It may all be exacerbated to some extent by the low power Continuous Descent Approach. However, unless the fuel quality control chaps in Beijing are being very circumspect about doing their job, I guess some uplifts may contain varying percentages of FSII (anti-icing additives and pump lubricants). Bowser-delivered fuel is another source of both contaminants and variable additives.
.
As you say,
<<"the reduced thermal capacity of the remaining fuel means it does warm significantly by the start of approach">>. Those waxy globules wouldn't be likely to emulsify back into solution; they'd just break down, as described, into smaller units and migrate downstream that much more readily (particularly once the demands of a throttle up starts sucking them through). Engines under acceleration demands have always been vulnerable to poorly adjusted AFRCU's or (so it would appear) to FADECs and EEC computers in receipt of outside-of-normal-parameter data.
.