Confirmed.
Bigio from Columbia to UCLA

Oh really? How about you back up those claims and tell us why the papers are weak?
Because I'm having a bad day, I'm going to take you up on this opportunity. For the record, I'm looking at the currently posted version of the paper, dated 2013: https://sites.google.com/site/jenlaoresearch/research
There are many levels on which Bigio and La'O is a messedup paper. For one thing, it bills itself as being about financial frictions, but all it has are static reducedform wedges on gross sectoral production that are isomorphic to conventional markups from monopolistic competition. Nothing in the paper really has to do with financial frictions in particular, until some halfhearted, unconvincing argument in section 4.3 that wedge movements reflect financial frictions, which uses a regression with RajanZingales industry dependence on external finance. The paper is mainly about how wedges aggregate in an inputoutput network.
The really terrible part is that, for a paper that's basically just about how wedges aggregate in an inputoutput network, it doesn't even get that part right. There are some really elementary errors. Take, for instance, Proposition 10 on page 35. This purports to show how real GDP depends on sectorlevel productivities and financial constraints, in a special case where labor is supplied inelastically (with some additional simplifying assumptions).
If you have good intuition for microeconomics, you immediately realize that something is wrong here: the expression for GDP is loglinear in the parameters "phi_i" reflecting the financial constraints of each sector i. This obviously can't be right: if labor is the only factor, and aggregate labor is fixed, then the undistorted equilibrium of the economy corresponds to the allocation of labor that gives us the maximum real GDP, and the distorted equilibrium of the economy will fall short of this maximum to the extent that wedges cause misallocation of labor. Locally, in the neighborhood of the undistorted equilibrium (which corresponds to phi_i=1 for all i), the shortfall will be second order; in particular, locally the (left) first derivative of real GDP with respect to log phi_i will be zero. Yet their Proposition 10 indicates a positive derivative, and indeed that GDP is globally loglinear in the phi_i. Clearly such loglinearity is impossible, and they've made a serious mistake.
If you look at the proof of Proposition 10 in the appendix, it's pretty easy to spot the mistake. On page 65, they bafflingly wrap up the proof by saying that "real value added in the economy" equals the real wage 'h'. Since the aggregate labor endowment is normalized to 1, this would be true if all income went to labor; but, of course, it doesn't, since the whole point of the financial friction is that it's associated with profits, as their equation (21) in the main body of the paper correctly sums up.
Proposition 10 is their only analytical result for the general inputoutput environment, so this seems like a pretty big problem. It certainly doesn't give me any confidence that they're doing the subsequent numerical analysis correctly.
So there. This is the example you wanted for "why the papers are weak". It only took me eyeballing the paper for about 10 minutes to find a huge mistake in the main analytical result. Geez guys, if you're going to write a trivial paper about how wedges aggregate in inputoutput networks, at least it right!

^ The undistorted equilibrium in their model with fixed aggregate labor solves a planner's problem, optimally allocating labor between sectors to maximize real GDP.
The fall in real GDP from any perturbation around this allocation is second order, simply because it's an interior optimum and a perturbation around any interior optimum in any problem will always have a secondorder effect on the objective.
This includes the perturbation caused by the financial friction parameters being slightly different from the ones that give you the undistorted optimum (phi_i=1).