Other reasons to retire can include the tread has outworn its function and a GTX membrane has failed. I had a pair of Speedcross that lasted me to 1500 km. I guess any shoe with real rubber in the tread might lose some springiness, much like old tyres should be retired, but I'm sceptical about modern plastics losing spring. If you use them long enough they could go out of fashion , like I have a good pair of Lahars with 12 mm drop but most of my shoes now are 4 to 6 mm drop.
Yes, base it on actuals, not targets. Daniels tables put tempo in the range between 12 km and 1/2 mara actuals, which in my case gives a 10 sec/km range. But like HA77 I prefer to do it on effort from HR. For a 5-mile tempo run based on your 5 km time (time/mi) a Daniels spreadsheet gives:
What Cougie said - even more so - the constantly changing angles of your foot plant on trails negates any notion of motion control. In any case, there is evidence that motion control shoes do nothing to change injury frequency. Maybe just try a trail shoe of the manufacturers you like, in the hope that the general volume, upper shape, last shape will be what you like.
I wouldn't have any snacks at all, especially if it's a snack to get energy that doesn't have too much energy! Also, I wouldn't look to snacks to ensure I were eating essential dietary items. Depending on the race length you will get through most of glycogen stores in muscles anyway (e.g., marathon). IMO most of coping with longer distances is down to natural ability and training. It could also be that you are not having sufficient rest between hard sessions or you are not tapering before a race. In short - just what Cougie said