What about when you buy NEW car, do you rely on manufacture est. mpg? or medicine?
There are 12 stages(I'll list 8 here) in designing new product:
1. Wish List spec+ features - usually come from marketing team
2. Design & protoype to wish list. reviews and approves by marketing
3. Beta build: minor adjustment usually improvment, build prototype near production level(where you get your MTF chart)
4. Reviews & collects feedbacks from trusted users(not public yet)
5. Final review + very minor chance(s), almost no chance in hardware
6. Pilot build: 99.5% production level, dealers get to play at this level
7. Production build
8. If build by 3rd party, price quote will be sent out to qualified contractors
If I can jump into your interesting discussion I'd change it a little bit making it better fit - in my opinion - to Canon.
1. Design and discover something in the lab
2. Patent it
3. Ask marketing what it's worth and if/when should go into production
4. Estimate production costs and give it to marketing so they could work with sales on profit margin
5. Put it into the box and place onto the shelve
6. After X years take the box off the shelve
7. Recalculate costs and price.
8. Add 25%-30% to the price.
9. Make announcement
10. Send box to the factory with a note "make me more of these but change the font and make the red ring more shiny"
There can be step 6.5 - Take another box off the shelve with similar contents inside and decide (or throw a coin) which box leave and which to take
The question remains on the decision process: which box and when should we take off the shelve? Maybe the cleaning lady in the warehouse decides looking on the dust amount?