Our your second point - the problem we are solving is human error and principal/agent problem, which can be addressed by management controls. We have seen the management controls fail when there varying skill and experience levels in the maitenance crew, when the operation has a transient workforce (say contractors at a harvest) or when the records are falsified. Falsification of records is principal agent - the person responsible for greasing doesn't wear the cost of downtime, labour and spares to fix it.
GreaseBoss can be used to eliminate all of these factors - customers who buy it may be a gold mine in PNG, or a potato farm in Victoria, a smelter in India. These types of operations have minmimal management controls compared to a modern developed world operation. As with everything, GreaseBoss has a sweet spot in the ecomony and it may not be where you have previously worked.
On your second point - we added a computer to a grease gun... that runs the risk of over complicating a simple tool, let alone an RFID on a zerk. We have worked really hard to make our system simple - we know our users are not going to tolerate a screen freeze, syncing errors etc - its not perfect now, but its on the trajectory. We have built our system with minmial impact to completing the current task, the only change is that you have to charge the unit at night. The only input interface is the RFID (no buttons) and the output on the screen. We want our users to pick it up and it just works.
With regards to the final comments, we have a version of the system that can be deployed locally - therefore no risk of losing data, bricking the device. The system works entirely offline, it only needs to sync once - provided the schedule never changes. And the third party can guide you around the plant, tell you how much grease and when its required, if you have a day off it can ensure the greasing is completed to the same standard as you would.
I hope this answwers your quesitons - thanks for making me work hard :)