What kind of propitiatory knowledge do they have that there are these long waiting lines for the machine?
What is they key issue that makes these boxes so hard to replicate?
Marketing (for example, BFL equipment is much prettier than Avalon).
What kind of proprietary knowledge do they have, that there are these long waiting lines for the machine?
None. But they do have capital, which they got by having a fairly good reputation from making cheap FPGA miners which they could do because they found a way to buy obsolete Altera FPGAs really cheap.
What is they key issue that makes these boxes so hard to replicate?
People who know what they're doing don't design Bitcoin mining equipment. There are probably a lot of different reasons why, mostly due to very high financial risk. The only people designing this stuff are incompetent noobs, with predictable results.
Why? Is it not profitable enough?
The capital they have is from taking pre-orders and not delivering. They've also burned a lot of that capital spamming the internet with ads to sell more pre-orders. If they ever had a good reputation they have squandered it by letting their loud-mouthed-pompous-ass spokesperson loose on the forums.
ICs are blazingly fast, because they have only one purpose - to compute SHA2 / to mine bitcoins. There is no overhead that processors & GPUs had.
A production of the IC costs high hundreds of thousands of dollars, and requires a ton of domain-specific knowledge. But once they're produced, each batch is quite inexpensive, and is capable of quite fast.
Of course there could be other startups doing the same thing, but you'd need an around a million dollar budget. Not easy.