So in the end I went along and just tried my suggested method. I let the primary cool down to 2 deg C (35F), made a priming solution and cooled this also. Then both where added to a secondary vessel to let it cold crash for a week. The temperature did increase a bit during transferring, to 4 deg C. The gravity with the priming solution was 1.016.
After a week I bottled directly from the secondary, from the tap on the bottom of the vessel. I moved the vessel from the fridge to the counter top and let is stand for a few hours. The gravity was still at 1.016 and the water lock had a ‘negative pressure’ suggesting there hasn’t be any fermentation (possible there had been some fermentation, but the CO2 will dissolve in the cold beer?). During bottling I’ve noticed that the first runnings where quite hazy. In the middle it was really clear but closer to the bottom it became hazy again. I marked the hazy ones to compare at a later stage.
Now, after a few weeks in the bottle, the carbonation seems fine, although a bit low. I like my Belgians foamy, so I usually aim around 3,0 vol of CO2, but compared to other beers it is not there (yet?). I took one of the hazy ones (also hazy in the glass), I want to save the clearer ones for better moments.
Conclusion. Was the experiment a success. Yes. It is possible to mix in the priming solution and let it cold crash at the same time. I agree that it is possible that some fermentation occurs, but due to the temperatures it will be minimal for most yeasts. Will I do it again? Probably not. Transferring and mixing beer just before bottling has a big plus and that is that it creates an even mixture of the yeast for each bottle. Right now I ended with lots of bottles with uneven amounts of yeast and therefore uneven clarity and probably conditioning times etc. For me, getting consistency across a batch weights more than increased infection risk etc. I need to check if one of the clear bottles really makes a difference but even if it does, I still have ~¼ of the batch not so clear.