You are indeed being too demanding, as can be seen when you know the general pattern of how the prediction was made.
Most industries, most of the time, show exponential improvement over time on whatever their primary measure of quality is. Moore's law is the most famous simply because its speed has had such dramatic effects, normally improvement is smaller, for example the 7% increase/year in battery capacity for a fixed price that Tesla's future business plans rely on.
See this graph of the cost of memory storage in hard drives: http://www.jcmit.com/disk2015.htm. (Note that an exponential curve is approximately a straight line on a log graph.)
Every so often there will be breakthroughs which show up as sudden jumps. After the breakthrough is assimilated you'll often see a pause because research plans were disrupted by the breakthrough, and then the curves start going again. Sometimes at a different speed. As an example on the above graph, the sudden order of magnitude price drop between 1995 and 2000 is due to the commercialization of Giant Magnetoresistance (GMR).
This pattern of technological advance is well-known, but the jumps are not predictable in advance. However even after such a jump, your best bet for projecting continued improvement over years to decades is the same methodology.