Obsolesence and device defects - I can imagine a saddle has a useful life of a couple of decades.
What are you really going to measure here beyond GPS and accelerometer data? The temperature and humidity of the saddle/horse interface? The strain and weight on the saddle? That'll allow you to report useful stuff like "your horse is tired and about to overheat", "you are to fat ride this horse", and the ever popular "this saddle is the wrong size for current horse/rider combo, please contact a sales representative". All the while taking in the views through the new and improved augmented reality interface on your smartphone, or hmd.
I mean why even have a horse at that point, it seems if you don't want to pick up on the communication from the live animal your riding, why not get a mud-bike.
Instead let's look at his product, which was his actual article goal, for people to think about and discuss the product. It's a sort of plastic or aluminum wine bottle that has a LCD touch screen on the side and wifi. We can see exactly where this concept came from, it's the IoT model. Put wifi on everything, and if its large enough, put a touch screen on it as well. Plus some sensors, like the temperature sensor and a clock that pretty much costs nothing to add since most embedded chips have a temperature sensor in there already. So for each thing in everything: add wifi and touchscreens. At some point thing is wine since wine is part of everything. Thus this product's provenance is in asking "What can we do by pairing a wine bottle with wifi?" And so they answer this and make a product to answer that pairing. The design process and reason for being has nothing whatsoever to do with what anyone wants or would be or that would sell or be useful. It's about pairing every conceivable preexisting product with wifi and touchscreens and forcing them to do something, anything, and desperately hoping it will somehow catch on, no matter whether it makes sense, no matter whether it addresses any real problem or need.