> A blockchain is just a database which broadcasts signed transactions over a p2p network.
That is certainly the definition some people seem to be using. However, that's not even remotely the innovative part of Bitcoin. If that definition of "blockchain" is acceptable, you might as well call git a blockchain.
What others consider to be the key innovation in Bitcoin is the combination into a self-feedback system of proof-of-work-mining (variable according to network power as determined by looking back at previous blocks), longest-chain-wins and the ever-growing-ledger with standard-rules-for-ledger-validity. The real genius is that this creates an incentive to mine while making double-spend impossible without 51% of the network (which is made difficult because others have an incentive to mine, and so it goes round).
The Wikipedia definition of "blockchain" appears to be a hashed chain of blocks (so basically git) with the addition of a selection-of-winning-branch algorithm. I'm not sure that's really generally accepted as the definition of "blockchain", though. The term seems to be suffering from inflation at the moment as everyone working on anything related to keeping a distributed ledger jumps onto the hype bandwagon.
This article is no better. It reduces the real innovation by referring to "known algorithms". Thus we don't even know what the author considers a definition of a blockchain to be!