Skype is famous enough to have reached the status of a verb, a rare honour. You don’t make a Net phone call, you ‘Skype’, just as you once you didn’t vacuum a floor, you hoovered.

The tiny problem is that, for all its brand success, Skype isn’t yet a giant at all, and certainly not in the league of Google. This might or might not explain why eBay, which coughed up $2.6 billion for the company two years ago, has this week sidelined Skype founder Niklas Zennstrom, and owned up to a degree of disappointment at having to take a $1.4 billion charge on the deal.

Poor eBay, such a lovable company too.

Some think the price was right, but the buyer wrong. Nokia should have bought Skype, not auctioneer of household dross, eBay. eBay thought it would help its core auction business using some hitherto undiscovered level of integration between the two, whereas everyone else knew that eBay is really just a giant junk shop on the Internet and that most people hang Skype phones of their computers to work around the over-priced charges of the BTs of the world.

The two overlap a bit, but the motivation for using one is not the same as the enthusiasm for the other.

But there is a deeper problem with Skype that few of the balance-sheet analysts have picked up on because they don’t spend time worrying about low-life protocols. Skype is welded to a proprietary, non-SIP (Session Initiation Protocol) based way of looking at the world, which turns its service into a sort of walled garden with no way in for outsiders other than the hidden door of its API.

The problems with this approach are unpredictable. Although Skype’s now notorious server crash of August would have brought down any VoIP or IM provider, the proprietary nature of the Skype enterprise looked like another way of hiding the network’s inner core and possible vulnerabilities behind a blanket of secrecy. For companies in particular, thanks to encryption, Skype is hard to see, is radically de-centralised in its workings, hard to block, often hard to manage, and this makes the risk posed by it hard to assess. Skype is a big unknown and that makes some people nervous.

Proprietary designs can be highly successful, but only when others are allowed into the fold to develop to the APIs. Skype has started to address some of these issues, licensing its API to IM control company FaceTime in an attempt to give companies a means of taking Skype’s tendency to go feral with sharp, pointy teeth, but it needs to go further than this.

People don’t download Skype because they want to use a fancy application or love Skype itself; they download it to make phone calls or to conference. Would the company be more successful, less successful, or about as successful had is started with SIP in the first place? It would almost certainly be somewhat less successful because the moribund telecoms companies that fear every new technology would have set out to detect and block it in their traffic. Now that it has won that battle, it faces a sterner challenge, that of increasing subscriber growth among mainstream companies without reinforcing its reputation for oddity and insecurity.

Companies fear what they can’t see and quantify, and companies can’t see or quantify Skype any more than can mystified parent eBay.