this is one of those oh-so-tricky sayings, meaning to take a small problem (the molehill) and make it seem much larger (the mountain).

fiendish, no?

Dimensional analysis would suggest that the critical angle for dirt piles would be independent of the amount of dirt (for amounts of dirt larger than, say, 10⁵ dirt particles). Assume the height of an average molehill is 15cm, and that you wish to build a 1.5km high mountain on the beach (i.e. starting from a flat base at sea level). Then you're looking at something 10⁴ times larger, but with the same shape. Therefore you'd need approximately 10¹² molehills to make this mountain.

IMPORTANT NOTES:

1. Most mountains exist in mountainous regions, therefore the assumption they stick up 1.5km higher than their surroundings is patently ridiculous. But even assuming we just wanted to go 150m above a plain, we'd need about 10⁹ molehills, which is still fairly large (see below).
2. The cost of earthworks is usually calculated by the volume of dirt you wish to move. However, in this case you're also lifting the dirt a long way up, which would increase the price (need to construct a road, pay more for fuel, and make longer journeys with each load).
3. Also, most construction cost calculations would ignore the cost of finding 10⁹ molehills in the first place; you'd also need to send out teams to dig them up and load them on your truck.
4. Mountains don't really look like molehills; they tend to have cliffs and huge rocks, which you cannot get in an accretion of molehills (some of the rocks will be far larger than even a complete molehill!). Therefore it's probably impossible; however, the same calculations may be performed for hills, which are probably more like molehills.