Can you post the full sets returned by the view for both tables?

The numbers look a bit odd.

Was there any DB shrinking done after the index was created?

And I would suggest at least reorganising the index pages.

But fragmentation should not be the thing to evaluate when comparing HEAPs with clustered tables.

Heap does not have a logical structure, so it cannot be fragmented, unless it's divided into several heaps.

Like a heap of rubble on a demolition site - it's a single continuos fragment, until you're trying to "cluster" it into chunks of bricks, concrete block, timber, etc. Then it becomes heavily fragmented, even though not a single piece of rubble has been moved.

What you should compare is the speed of selection.