On a phone by phone basis, yes, but I imagine that they used sales forecasts to determine that the extra revenue per phone sold with 32GB would be outweighed by the increase in the conversion rate for upgrades to the second tier of storage if that tier included 64GB.
My guess is that so few people would buy an 8GB iPhone that Apple would actually lose money on the model (relative to a lineup including a 16GB base model) due to having to buy lower quantities of that memory chip at higher prices and having to make small production runs of that model. There's a threshold below which it becomes uneconomical to produce the variety of styles the iPhone offers because of a higher risk of inventory obsolescence (e.g. what if nobody buys the...
You must have missed the first sentence of the post you quoted. I'll repeat it here since it answers your question.What Apple knows is that people are more likely to upgrade to the second tier of storage if the first is inadequate.
You're misconstruing the reason that the 16GB model is selling quickly. It's because it's the cheapest model, not because it's the favored storage capacity. People buying that model are either deciding that they can find a way to make do with 16GB or they don't have any comprehension of what that capacity means in real world usage.
What Apple knows is that people are more likely to upgrade to the second tier of storage if the first is inadequate. There are a lot more people who are comfortable with 32GB than 16GB, and that would reduce their conversion rate on upgrades.
Obviously 16GB of storage space is very limiting regardless of how much the phone costs. The difference is that a $380 phone can be expected to have compromises. A $650 phone should not, let alone one costing $750.
I wouldn't expect the training to totally eliminate that problem. You have to train Android before it will allow OK Google to activate the search interface, and as you've discovered, search can still be triggered by other people if you're able to imitate the phone owner's speech at a basic level (pitch, cadence, etc.).