I think there is less here than meets the eye. We will hear about signalling because the new big seed rounds have added another option to the collection of plausible sounding excuses VCs use when they don't want to invest. But as with the other excuses, it will always be less of a problem in reality than it seems to be from what the VCs are saying.
If you judged by what VCs say when they turn people down, it would seem like the most dangerous thing you could do as a startup would be to start a company VCs couldn't add much value to.
This is an important thing to consider. It used to annoy me that people would dismiss an idea or company because of some peripheral (and not necessarily material) bit of information. Jim Breyer was an investor in FreeGate (a startup I did in the first bubble) and he explained it fairly simply, he said "Chuck, people invest (or don't) for their own reasons, not all of them (either the people or the reasons!) are rational." And what that means for the budding entrepreneur is that you have to cater to the irrational ideas that live in the ecosystem, even though they are irrational. The whole "we won't invest if your seed round investor isn't" is similar to people who won't hire engineers if a company that they did an internship with didn't offer them a job. Irrational, but it's something to be aware of.
I think it's also due in large part to VCs giving you one reason when it's actually another. They want to leave their options open. Sometimes saying the real reason would leave you not wanting to work with them in the future. Giving you a trivial excuse that seems stupid is better than one that's offensive or makes them look like an asshole.
another reason to avoid VCs in the seed round - those VCs will know they have leverage due to signaling issues and offer worse terms. That matters when you're selling 20%+ of your company.
Angels are on your team when they have 50k in you, but VCs are not. They are not truly om your side until they have a few million in you.
In a somewhat related note, I noticed that A16Z invested in Instagram in the seed round, but is nowhere to be seen in the later rounds (http://www.crunchbase.com/company/instagram). Wonder what happened there?
A16Z invested in PicPlz, which was a competitor to Instagram, after the initial investment in Instagram (which was then called Burbn and was doing something else). So, after Burbn pivoted to Instagram, A16Z had to choose between 2 competing products, and they chose PicPlz for further investments. Not an inspired choice, but they still took a good chunk of money from Instagram's $1B acquisition.
It does apply to later rounds, but the situation in the seed round is much different. In a seed round, you have the option of raising from either angels or VCs, including potentially many different VC's. This gives you some control of the situation and changes the signaling risks as compared to raising a series B.
Even more importantly, as Elad mentioned a VC firm can do many, many seed investments per year without much cost to them, but is very limited in the number of Series A investments. The rate-limiting resource here is time more than money; a VC can only sit on so many boards. While this makes a Series A investor passing a worse signal than a seed investor passing, the former is in practice much less likely to happen because of how resource constraints work.
Indeed, signalling concerns apply to every round, but perhaps they are more important for the seed-to-A transition than for later rounds.
It seems that most companies that raise a seed round (and survive) go on to raise a series A before an exit, but many companies that raise a series A can get to an exit without raising subsequent rounds. (Admittedly, I don't have solid numbers; this is just based on casual observation) Thus, future signalling worries could be less important during the A round if you aren't as concerned about ever having a B round.
Agreed - the transition is really seed to series A.
By the time you hit series B or C, you have investors who own a large part of your company already (and in most cases won't do an "inside round" - or if they want to do it, the entrepreneur does not).
So the dynamic post-series A becomes the VC helping you more in raising a B, but not competing for the series B. Since they are not competing to lead your B (they just pro rata), there is no signal (unless they do not want to pro rata, which is a very large negative signal).
If you judged by what VCs say when they turn people down, it would seem like the most dangerous thing you could do as a startup would be to start a company VCs couldn't add much value to.