Gresham's law says that "bad money drives out good". Due to the dynamic this describes, it is unlikely bitcoin would commonly be used to buy and sell things even if it did not suffer practical obstacles.
Bitcoin experiences less inflation than regular currencies. Some coins get created now, but over time we know its character will become deflationary: no new coins will be created, and some will be lost at times due to poor wallet management.
As a result, people will chose to spend other currency in preference to spending bitcoin. This is self-reinforcing. The infrastructure will not be in place to use it on the odd occasion that someone wanted to.
You could create a blockchain currency which had a natural and continuous rate of inflation, to encourage people to spend it. You could bootstrap this by mutualising it across an industry. e.g. imagine if the largest datacentre groups got together to create ModestlyInflationaryCoin, and then said they would offer discounts to customers who paid in ModestlyInflationaryCoin, as a means of bootstrapping it. Other groups might start to use it, and it would stay in circulation because people would want to be rid of it once they had it.
Even if such a currency existed, it would probably be short-lived. /Once it was bootstrapped/, its stakeholders would have strong incentive to change its contract to be non-inflationary. Making that change would convert their holdings from Bad Money to Good Money, and as a result the character it would significantly increase its value.
But the datacentre groups could then mutualise a new currency in place of the old one. It is possible that there is a virtuous loop here, and that there will be a race to quality in currencies in our future, grown from how easy it is to create new currencies. We might start to see the identity of currencies a bit more like the way we see futures contracts in our current era.