Blockchains are upgraded form of databases. <…>
No, the opening statement is not true. In fact, if it were to be truly compared to a conventional database (say a RDBMS – Relational Database Management Sysyem), blockchain loses on many fronts and therefore cannot be considered an upgrade in any way.
Blockchain is slow when it comes to storing the data it holds, and the algorithm’s used to confirm transactions are nothing like what a plain RDBMS commit looks like. Blockchain also requires incentives for the TXs to be confirmed (imagine having to incentive our databases to get the commit done). On the other hand, the distributed and replicated nature of blockchain make it more tolerant to hardware faults and software attacks such as virus or randsomware attacks.
The core goal is really different from the start: blockchain acts as a public ledger, where movements are logged along with some related information (quantities, origin, destiny – be it with the degree of anonymity the specific blockchain provides). Databases will hold data of all sorts of nature, even complex structures such as blobs and objects. This degree of heterogeneity is not the intent of the blockchain, and the speed at which on and another processed the information is radically different.
We can compare features of one another, but really we’re talking about two different things in essence, with different objectives and features, making one not an upgrade of the other in any way.