Are Institutions Insignificant?

Or in other words, is President Aquino's emphasis on "kung walang corrupt, walang mahirap" overstated? Francis Fukuyama weighs in:

Over the past decade the mantra in both development studies and comparative politics has been “institutions matter”—that is, you aren’t going to get economic growth or other human development objectives in the absence of institutions like rule of law, transparent and accountable governments, low levels of corruption, and the like. 

The empirical basis for this assertion is actually much weaker than many of us would like to think, however. Plenty of countries, beginning with China, have grown very rapidly over the past generation in the absence of what is now called “good governance.” Indeed, the US and Britain charted the industrial revolution with governments that were substantially more corrupt and less capable than they are today. 

...The Bank of England became independent only in 1998; there is no British constitutional court and therefore no judicial checks on legislative power; not just 2/3s but a fifty percent plus one majority in the House of Commons is sufficient to overturn any law in the land, including any protecting England’s fabled press freedoms.

Whenever I read articles like this, I wonder -- what exactly do people mean when they say "institutions"? Do we know what it is? How do we measure it?

Perhaps the very reason there is scant empirical evidence on the effect of good institutions on development is because it is simply poorly defined, or people just can't agree on what it means and therefore talk over each other's heads. Fukuyama here rebuts by presenting himself, scant empirical evidence.

I do suspect sometimes that we what we mean by "institutions" is simply everything else that we can't explain or understand -- that is, the residual. Any thoughts on this from someone more knowledgeable?