A MimbleWimble design is good for pruning. If you are familiar with UTXO model in Bitcoin this can be explained in comparison. A transaction in Grin contains outputs and a kernel which is a signature. Spent outputs can be pruned from the blockchain. Kernels cannot be prunned, but all the transaction inputs are signed with one signature. With all above, grin should use storage more efficiently than bitcoin. Don’t know exactly, maybe 3-5 or more times depending on actual utilization.

Currently the grin node uses about 1GB with pruned blockchain, it is 1/3 of all the historic data: https://grin.blockscan.com/chartsync/syncarchive

I personally don’t like the bitcoin emission curve. Its a curve of looser coin. Satoshi never thought about it success and created a niche currency for cypherpunks. For a niche currency which adoption doesn’t grow over time fixed supply is fine, but not for a world currency. Bitcoin adoption will be hard and painful. Fixed supply leads to a shitty hodl culture and cryptowinters.

Monero is good with it’s privacy features. Also, they have some good features inherited from the bytecoin design: smooth emission curve, dynamic blocksize limit. But it has disadvantages anyway:

• Transactions size is ten times larger than in bitcoin. If monero has as much txs as in bitcoin its blockchain will growth ten times larger and/or fees will be significally higher;
• Lack of infrastructure and real world use. Breaking changes happen often and prevent from integrations in apps;
• No active develop or research of Lighning Network/payment channels or other scalling solutions like sharding. Lack of activity makes onchain txs enough. LN cannot be activated in one moment by pushing a button. It needs channel infrastructure, routing network and apps to be created;
• ASIC resistance PoW algo makes 51% attack easier. With a block reward reduction to it’s minimum in 2022 attack will be even cheaper;
• Benefits for early participation with the emission curve make it harder to get adoption.