I'm guessing it's more to do with transaction times than anything. A merchant needs to wait 10, 20 or 30 minutes to get a confirmation in Bitcoin, because he can't be sure there will not be a double-spend of the money until it is confirmed.
Technically, this has always been the case, but there was a false belief that 0conf transactions were secure, because Bitcoin clients rejected double-spend attempts by default (as a DoS protection). This prevention of double-spending unconfirmed transactions was never part of the protocol though, and it was made clear when Replace-by-fee was added, which shows how clients can easily replace existing transactions with new ones, without changing the Bitcoin protocol.
There's still a crowd who believe that 0conf transactions are fine, and RBF should be removed (eg, the Bitcoin Cash developers), but it's really just ignoring the reality that RBF is not an enforcible protocol rule, but a client-side policy in the software. It's probably fine to use 0conf for small transactions because it is manageable risk, similar to accepting credit cards where payments can be reversed.
Another possible solution to faster payments is with payment channels, where the risk of double-spending is mitigated by hashed time-locked contracts which are transmitted between parties and not broadcast over the Bitcoin network until one of the parties wants to reclaim their funds onto the Bitcoin network. Since the parties are just exchanging HTLCs over a TCP connection, they're practically instant, the fees are low or nil, and there is a potential privacy improvement by not revealing every transaction into the Bitcoin ledger. This is in development (minimally working on Lightning Network now), but proceeding quite slowly.
If Bitcoin is going to have widespread global usage, these problems need to be overcome so that it is as simple, or simpler to use than credit cards for the average user. It possibly is simpler to set up for merchants already, if they're technical enough, because they don't have to involve 3rd party payment processors.
The other side to it is that we take credit cards for granted in the developed world, and often miss that there are a billion or more people without banks in the developing world. These are the people who have the real use-case for Bitcoin. A single user can set up a Bitcoin node on a smartphone in some remote village, and suddenly bring banking capability to their entire village. I think that's a very strong use case. People are also using it to hedge against their national currencies which are being hyper-inflated by their own governments - as is happening in Venezuela right now.