CRV loans up to $250k which opens up a wider range of start-up possibilities. E.g. if you are working on something involving hardware; remember Apple really got moving with a $250k line of credit backed by Mike Markkula (consumer price index says $250k in 1977 is about $855k in 2006 dollars).
Though frankly, Octopart isn't exactly a "simple website" either. It's a search engine, and in some ways more complicated than Google. Is Google a "simple website?" (If so, bring 'em on.)
For example, use the suggested "red led" search and compare octopart to newark. With newark you select criteria that apply to the part; with octopart you "keep quoting" "keywords" "to:narrow" "it down" and you can't select a range -- like a range of voltage, for example. I'm not sure if you can even "or" a few keywords together to get a range.
There's nothing wrong with it, I'm just saying it isn't that complex. Reddit is cloneable in a weekend (remember all the "reddit in 20 lines of Lisp" posts after it switched to Python) and most likely owes its popularity to you personally; it had a good foundation because its earliest users were your readers.
Obviously it was a smart investment. I'm just saying it wasn't a hard technical problem. The barrier to getting a large (smart) audience was solved by having a PG to promote it. You're (personally) essentially a catalyst, like start-up yeast.
If you re-read the OP, I said YC has a bias towards a certain type of start-up, not that it was exclusive. If you list all the YC'dlings it may hold true for the majority.
As for the exceptions you mentioned, Xobni hasn't released anything, nor has the unlaunched unnamed start-up, so I'll have to take your word for it. I don't have Boost so loopt isn't something I can even use yet. As for TextPayMe, it reminds me of the pre-web Paypal, and a natural extension of what Paypal's doing now: adding an SMS option. (Perhaps less clunky than using your phone's web browser to go to paypal.com.) However, sending and receiving text messages is not exactly rocket science and the barrier there is really in handling the money and dealing with the applicable regulations, then having a large enough network to be useful. Paypal already has the hard part done, and I can think of numerous reasons to add SMS capability in-house instead, but eBay has so much cash they might just decide to acquire you. (I'm personally impressed with TPM's gumption in dealing with what I imagine is quite a lot of red tape.)
Was implementing Reddit harder than I think? I think the original route was circuitous, fighting a buggy CMUCL and lack of web frameworks.