I suspect, Branson's Bitcoin investment is very limited. From what I understand most of those accepting Bitcoin for commerce are doing so with a lot of caution and on a very limited basis.
There is little risk IF you almost immediately sell them. Then Branson's loss is very likely, if any, to be less than when his customer pays for his ticket with a charge card.
Next day by edit (after accidental discovery that You can get live bitcoin quotes at:
http://www.kitco.com/finance/bitcoin/ ,which seems to be an on-line market too.):
The buy sell spread at 8:37am NYC time on 7 Jan 14, was: bid 892.16 / ask 895.13 (or $2.97). Thus currently the worst cost of a quick re-sale is a 0.33% loss, or nearly 10 times less than accepting a credit card.
Branson is no dummy - probably hopes all his customers pay with bitcoin instead of their credit cards.
The real cost to the buy and hold bitcoin speculation is the "opportunity cost." You pay up front and lose potential interest on your money. Also the value seems to go down as easily as it goes up. Bitcoin make a lot of sense for merchants, like Branson, as effectively most of their cost (that gives profits to credit card companies) can be transferred to their customers. It is the government and credit card companies who are the sure losers, if Bitcoin is widely used with current rules. Governments have power and don't like to lose, so expect to pay taxes on any bitcoin capital gains and perhaps other fees, like toll for use etc. (A monetary toll bridge of 2% on every transaction?) If there is any way for IRS to get data on all bitcoin transactions, a monetary toll bridge is a gift to the IRS and sort of a "national sales tax." IRS would not need to know who buyer or seller were or what goods were traded - no violation of privacy - Just take 1 or 2% cut in bitcoin as they pass thru the clearing houses between buyer & seller. As I understand it that data is already public, just the ID of the traders is not. As they say: "The road to hell is paved with good intentions." (Or should that be: "paved with bitcoins."?)
In my opinion, gold, is much better bet now than bitcoin. Gold has only gone up in 2014 - as I predicted it would - See:
http://www.sciforums.com/showthread.php?134352-Gold-Bubble-goes-POP&p=3150435&viewfull=1#post3150435 and early posts that tell why (Comex vault lost 80+% of the deliverable gold it held in 2013, mainly to deliver to Asians real, not paper, gold paying 20+/- 5% premiums for the "real yellow stuff.") Even if there is no accelattion in the rate of people with long contracts climing physical gold rather than accepting dollars, Comex vault will be empty in about 85 days (assuming Comex does not start buying and
paying what ever is demanded to avoid defaulting on delivery contracts.)
Of course there will be acceleration as now ~80 paper gold traders have a claim on the same ounce of gold. Some already are asking for their gold before one of the other 79 who can does. (Only a couple of months ago that claims to gold ratio was 50.) Most will get "German like" promises that they can get real gold in 2020.
I'll go further out on a limb and predict that on or before this Friday gold trades, in US, at $1248/oz. Note that is a non-convergent geometric series start. (It is already doing more than that in China!) Anyone with the gut to bet it does not?
later by edit: I just checked at Bloomberg (11:34 NYC time) and gold was already up today $3 to 1241.60/ oz. One still needs to wait until mid January (end of my first prediction of rapid price rise) but it does seem like the Paper Gold traders are losing control of the price of gold to the law of supply (too little) and demand (soaring beyond total gold production, including the re-melting of jewelry).
By later edit: at 2:50PM NYC time, Bloomberg gold was up only a dollar at $1239.60 continuing the pattern I have notice on most (all ?) 2014 days. I.e. the peak of the day is just after normal bed time in East coast cities of China and then falls down some. - May be a diurnal change in Chinese demand, doing this?