This question might be considered too broad and too opinion based according to the rules of this forum. However, tackling it from a technical standpoint based on facts rather than opinions might be on-topic and I will try my best to deep dive on that. Let's try to hit the five most important aspects that people might be interested in: security, decentralization, scalability, governance and privacy.
The most important security consideration that needs to be given to a digital currency is preventing double spend (spending the same coin more than once), fraudulent transactions (spending what is not yours) or creating money out of thin air. Bitcoin prevents all three of this by storing its transactions in a blockchain that are almost immutable after
certain blocks due to the proof of work (PoW) algorithm. Supply is capped, coins can only be generated by the mining process, transactions are governed by cryptography and coins can only be spent by providing a valid signature. The PoW is also an important security consideration for preventing double spending that is sometimes overlooked by many. It signifies the inherent cost involved to modify blocks and also means that the malicious party amasses more mining power than rest of the network combined. After a certain number of blocks the transactions becomes extremely difficult to alter which means the they are immutable.
Libra operates on LibraBFT consensus, a class of the classical Byzantine Fault Tolerance (BFT) model. A model that operates on BFT can be trusted as long as there are at least 2/3rd honest nodes operating the network. So 33% of the nodes can come together and launch a malicious attack. They also need not spend money in terms of hash power like PoW in order to do that. So collusion could be an easier possibility with BFT models.
Bitcoin is one of the largest distributed network on this planet. Entire data is public to all the participants of the network and there are no gatekeepers or requirements to become part of the Bitcoin network. Every full node operating the bitcoin software validates all transactions in its entirety from the genesis block. When miners mine a block, every full node ensures that the block is generated according to the consensus protocol and rejects blocks which do not meet them. This ensures full decentralization, and ensures that power is not concentrated in the hands of the miners as claimed by many ignorant people. Try reading about the block scaling debate of 2017 and you will understand how the community played a major role in governing the network.
In Libra's current implementation, 'trusted' validator nodes are core to the security model. The Libra blockchain will be open to the public to read, but closed for writes.
This means that hard power is centralized around these few entities and the protocol itself depends on these entities to validate the protocol. This means the protocol depends on whatever they decide it is. These entities can freeze your coins, take your coins, issue new coins, or really whatever they want - the sky is the limit. Libra claims that in the future it might move to a permissionless model, but it is highly doubtful that such a scheme can succeed.
Bitcoin was designed for transactions to be as secure as possible and that comes at a cost of sacrificing scalability for security and decentralization. However, the current Lightning Network implementation solves this issue and ensure that bitcoins can be transferred almost as fast as you can create and transmit them without sacrificing on the underlying security mechanisms.
Libra being 'permissioned' inherently has higher scalability. It is no different than doing a paypal transaction or spending via your credit card.
Trustlessness is the most important characteristics of Bitcoin. Bitcoin governance emerges from users through the software they run on their computers. Nobody calls the shots. The entire Bitcoin code is public and you can go and verify the entire program yourself. The governance components also involve the decentralization and the security considerations that have been mentioned above. Also, Bitcoin developers have no special powers to make users switch to a new network. Even if the Bitcoin Core developers were to release a new version of the software to create such a new network, users of the older software implementations could simply ignore the update and continue using the existing protocol as they please. That is the power of decentralization.
Libra being a stable coin, one needs to ensure that for every coin that issued, balances of fiat currency is maintained. So there involves trust in the organization to keep the balance in fiat currency, trust in the fiat currency itself and the instruments (or banks) in which this balance is actually kept. Also, given the centralization involved in Libra, the sky is the limit in terms of what the 'trusted' validators think the protocol should be. Most of these validators are Fortune 500 companies, and they are not there for charity. Their main goal/incentive lies in improving shareholder value and one might now be incorrect to think they can potentially go to any extent to achieve that based on historical precedence.
Bitcoin provides pseudo-anonymity. Although, many exchanges ask for KYC while buying bitcoins, the protocol itself does not need any verification. Even after buying from an exchange, you can do a CoinJoin transaction to hide your identity. Preventing address re-use and keeping keys under one's control will then ensure your privacy is protected.
At least from what I have read, the resellers of Libra will have to collect identifying data from users under current FATF recommendation if the amounts being transferred exceed $1,000. This data has to be managed and transmitted out of band from the blockchain. A lot of internet companies have come under flack recently for in-apt data sharing practices. If one were to draw a parallel, this data can overwhelmingly be used for many other business enhancing practices as well.