The important part is how blocks get added over network, and decentralized consensus of the addition. (Well, socioeconomic phenomenon of adding monetary value to all that jazz is pretty cool too!)
But yes, dynamic-membership multi-party signature schemes and their use for creating permissionless systems is the real innovation.
Do you (or anyone else) have a technical reference for differences between the Bitcoin and Ethereum blockchain length function?
https://bitcoin.org/bitcoin.pdf
Ethereum uses a variant of the GHOST algorithm, which was introduced in the paper Accelerating Bitcoin’s Transaction Processing: Fast Money Grows on Trees, Not Chains.
https://eprint.iacr.org/2013/881.pdf
The Ethereum white paper mentions how they modified it:
https://github.com/ethereum/wiki/wiki/White-Paper#modified-g...
Full detail is in their infamously difficult Yellow Paper:
I had some idea about the Bitcoin blockchain - also succinctly summarized by @tylerswift.
But the Ethereum material is all new to me. It at least sounds more interesting than "the least interesting".
It's still all standard bachelor's degree algorithms, and I agree with the gp that it's not the interesting part. The interesting part is using these algorithms for social consensus on value.
It's not that special, each invoice gets assigned as hash of (invoice number + total amount + date of emission + hash of previous invoice). Besides having to send that XML file to the IRS each month with all the issued invoices, the hash also gets printed on the invoice given to the client (paper or PDF), so it gets tricky to change the past without getting caught.