This world is full of equipment that has modular connectors. These modular connectors also exists on analogue phones. Joy, of joy, if someone connects the equipment to a phone line and the phone then rings. So is the equipment then expected to survive this ring voltage?

What I tried to show is that there are an almost infinite list of potential problems. Problems that doesn't feel natural to think about for the customer until they have received units and have sent them out to their end users.

A customer is likely to think it irrelevant with a question if a unit must be possible to operate in strong or weak light or by user who is colorblind - until their end users calls back and are irritated. How fun is it to use a mobile phone in very strong sunlight? If not voice-operated, it may be almost impossible to make use of the phone book to call someone whose number you haven't in your head.

A customer is likely to have an idea about how many hours a unit should be able to operate on batteries. But they seldom think about how many years for service maintainance cycles. Or highest/lowest temperature. Or how long the unit must operate on batteries if it is -30C. Or how fast the batteries must be possible to recharge. Or how the recharge speed can affect the life span of the batteries. Or if the end users will be able to find alternative batteries locally. Or if the unit should have a standard charger that can be field-replaced by visiting a larger store, or a custom charger only available for replacement from the distributor. Or implications of battery size in relation to air shipping of individual units or bulk shipping of units from the factory by plane. Or if there is a need for extra mechanical protection of the battery.

A number of open issues can be seen a clear features that you can check off - and where you can request more money if the customer wants a change. But for a number of situations, it can instead be seen as just workmanship. Do the customer wants a cheap device or a mil-standard device. And is the developed product of the quality standard that the customer expected? It's quite common that the specifications does not contain enough information about testing to let developer and customer do an independent test of the quality of the product.

Many products are of so small size that it isn't economical to produce a 500 page document regulating every tiny detail. The cost of ceating the document, and reviewing it and signing it would be higher than the cost of developing the product and include time/cost for one or two redesign spins.

Just think about it. How many times must the power connector support being removed and inserted? What physical force must it support if the user drops the device, but catches it in the power coord? Should the power plug manage to catch the device after a 1 meter fall, or should the plug release and the unit hit the floor? What resistance is needed to UV light, or chemicals? More than one product have managed to get the display destroyed because someone have stored the device together with some cables or other soft plastic, and the softener in the plastic have reacted and partially melted the display plastic cover. How do you remember to put such requirements in the document the customer must sign before you start developing?

An engineer have an almost infinite list of details to think about.
A customer only wants something that works well enough.
But what is "well enough"? And is it even possible to create such a specification?

A customer thinks function - until something break. First then, do they want to discuss limitations. Before that, they just assume the engineers should be skilled enough, while ignoring the cost difference between an average product and a extremely robust product.