This comes, of course, just days after I revisited the whole “VC model is broken” discussion from several years back, in the wake of Sevin Rosen basically returning its entire eighth fund (on paper) from the successful Splunk IPO. So a lot of you asked for me to chime in on the Kauffman study. Or, in the words of one reader, “admit that you don’t know what you’re talking about.”

The study basically breaks down into a pair of primary arguments:

1. Venture capital has, in general, not produced its promised returns for more than a decade (with the situation being particularly acute in funds >$500m). In fact, it finds that LPs likely would have been better off investing in small-cap public equity index funds.
2. Limited partners must internalize point #1, and make major structural changes to VC fund agreements in order to minimize their risk.

The first point is mostly based on data from Kauffman’s own portfolio of nearly 100 venture capital partnerships. And it is fairly compelling – particularly its argument that using vintage year quartiles is inadequate, considering that firms often fundraise at or near “peak IRR” (which sometimes comes much earlier than you might imagine).

Unfortunately, Kauffman does not disclose its GP relationships – leading to some charges from the outside that the organization simply picked lousy funds and is looking for a macro scapegoat.

Fortunately, I did some digging: Kauffman’s “core” venture capital portfolio includes multiple commitments with Bessemer Venture Partners, Benchmark Capital, General Atlantic and Madrona Venture Partners. It also has legacy relationships with such firms as Clarus Ventures, Mayfield, MPM Capital, New Enterprise Associates and Oak Investment Partners. From what I understand, no relationships with “brand names” like Accel Partners, Kleiner Perkins, Sequoia, First Round, Foundry or Union Square.

At first glance, my reaction was “Wow, Kauffman has such lousy returns with those funds included in the data set?” Either the other funds are god-awful, or the included brands aren’t the superstars they get made out to be (particularly when you learn Kauffman has never had a $1 billion+ fund return 2x capital after fees).

Then, a third option: Kauffman did a lousy job timing the secondary market. At the end of 2007, I’m told that the Foundation agreed to sell approximately 1/3 of its portfolio at 97 cents on the dollar (deal actually closed in middle of 2008). That was well done. Unfortunately, it continued selling after that as secondary discounts got deeper and deeper. For example, it dumped its NEA positions for around 40 cents on the dollar – which doesn’t look too hot in retrospect. More importantly, Kauffman only has data for these legacy funds as of the time of sale (i.e., a market bottom that came before the recent IPO resurgence, or the big value creation of an NEA portfolio company like Groupon or Fusion-io). To be clear, I’m not disparaging the Kauffman researchers – I’m simply wondering how representative its data set is to current conditions.

All of this said, however, Kauffman is correct in arguing that, for many LPs, venture capital may not work best as a dedicated asset class. The industry stratification is severe, and becoming self-perpetuating as the best entrepreneurs seek out the “best” venture capitalists. If you can’t get into a few dozen of the best GPs – including emerging managers — perhaps the next-best option is a public index rather than a second-tier fund (and by second-tier, I probably mean “below top-decile”).

Kauffman is also on target in its second point, about the ridiculous fund structures vis-à-vis GP/LP alignment. As it asks, would VCs ever give portfolio company CEOs the same deal they ask for from limited partners? (hint: the answer is “no way in hell).

For example, LPs should have information about underlying fund economics. If two senior partners get most of the carry in a firm with young stars, then that is pertinent information LPs should have when making a decision. More importantly, 2% flat management fees – as opposed to budget-based fees – create a perverse incentive to keep raising larger and larger funds, even if not merited by the market opportunity (and, remember, Kauffman believes larger funds underperform). And how come VCs put liquidation preferences on their own transaction, but generally oppose hurdle rates or waterfalls on their LP agreements?

But, remember, Kauffman doesn’t really blame greedy GPs for the misalignment. It blames LPs – the report’s title is “We have met the enemy, and he is us” – for refusing to advocate for, or accept, structural innovations. Mainly it believes that investment committees are either too unsophisticated or too set in their ways to minimize their own VC risk:

“One experienced GP raising his own first-time fund said he offered a budget-based management fee, and was open to a sliding carry based on performance (e.g., 25 percent above 2x), but felt that such a departure from industry practice sent the message to prospective LPs that the fund was desperate to attract new investors by offering unique and better-aligned economic terms.”

Or take an example from the buyout world, when Bain Capital recently offered investors in its Asia fund the option of 1/30 or 2/20. About half took each deal, with many of the latter explaining that investment committees don’t blink at 2/20 but ask lots of questions about 1/30.

There is, however, a bit of irony in Kauffman being the one to push for such things. Several years ago, the Foundation dropped out of membership in ILPA – perhaps the only organization with enough collective clout to actually effectuate change. The idea of individual LPs asking for budget-base fees or underlying economics – let alone broader structural alterations – is daunting. It seems, to me, that Kauffman has a willing partner in its fight, if only it would mend some fences.

And, again, risk mitigation only would have a limited impact on ultimate returns. The key remains picking the best managers. You can pay 0.5% management fees, but if the VC picks bad companies (or overpays) then you’re still in the red.

For me, Kauffman points out an important distinction in the “VC model is broken” argument, by delving down into LP commitments rather than GP commitments. I’ll happily concede that the former needs vast improvement, while continuing to believe that such alterations only matter because actual venture capital investments (i.e., GPs investing in private startups) continues to hold great potential for outsized returns. Sevin Rosen was arguing on the latter slope years ago, given that it killed fundraising altogether, rather than offering up a unique fund structure.

Early-stage investing remains fine – it’s what supports such investing that needs fixing.

Sign up for Dan’s daily email newsletter on deals and deal-makers: GetTermSheet.com