The only case where having the first 150k guaranteed would make a difference would be if we were considering a group that we liked, but we thought other investors would be too timid to fund. E.g. a group doing something insanely risky. But in those cases we always fund them anyway, and count on our own ability to talk the later stage investors around.
The companies that communicate a good idea well on Demo Day will succeed in getting good funding.
The difference between $20k and $150k is "They need to work on communicating their idea" vs. "They can get to work on their implementation".
If you are going head to head with something like ec2, $170K is probably enough to get you started, I'd think, if you are willing to live cheaply and don't screw up too much. If you count all the contracting money I put in, eh, that's probably in the ballpark, and I am to the point where I can reasonably provide compute units at a competitive price, though my software lags by quite a lot. I mean, you need /some/ scale, though you don't really need as much scale as they say.
Of course, if you are a high-margin infrastructure company, and your margins are high enough to re-sell ec2 with a value add software stack, $20K very well might be enough. I'm just saying, $20K doesn't get a lot of hardware
Not if you're making hardware...well, not all hardware, anyway. Was Wakemate able to get off the ground with 20k?
For example, maybe there's a startup which couldn't possibly succeed without big initial investment - with a <$20k fund, you wouldn't bet on them doing well, but with an additional $150k available, you might see them as having a much higher chance of success.
Sure there are other examples, but basically the fact that it changes the amount of funding available to your startups, it has the potential to change their success, which means that, if you forsee those changes before deciding whether to accept or reject a startup, it could also change your decision.