There can't be 2 blocks left in the memepool.
It only means more people used smaller fees in those blocks that's why the fee rate is low.
As you can see from the image only two blocks were left that day. Don't worry when I create the thread in the main board you will see the answers other people gave me in the PM discussion. And the thing happened on Sunday. I just want to know more from here.
Blocks are not found in the mempool, only transactions can be found in the mempool. What you are seeing right now is representation of transactions arrange according to their fee ranges. If someone see the mempool to be that empty and pay high fees, you will see the arrange blocks will reposition the fees again and it happens continuously when there are more transactions with high prioritize or low fees.
NB: There is difference between mempool as transaction explorer and Bitcoin node mempool.
Mempool.space is a Bitcoin explorer where you can scan and view different transaction ID while Bitcoin mempool is where transactions that are waiting for comfirmations stays before they get included in the the block and there is not such thing like general mempool, mempool.space doesn't represent the entire Bitcoin mempool, each Bitcoin node has their own respective mempool(memory pool) of transactions.
So where can we found Bitcoin Blocks since those blocks in the mempool are not in the bitcoin blocks where do we see them. Please drop link. But I agreed with you that after the two blocks has been mined, other blocks were generated so it is a continuous process. The transactions are not much in that day the mempool can be exhausted which happened on Sunday but they must be new generated blocks to replace the last mined blocks.
Make I do small elaboration.
Mempool.space na Bitcoin transaction explorer like them
https://blockchair.com,
https://blockstream.info,
https://bitcoinexplorer.org and many more but what made mempool explorer unique is because of their visual representation of transactions in blocks according to their fee structure in sat/vbytes.
When you send a Bitcoin transactions, the node are the ones that pick your transaction waiting for miners to comfirm them and this transaction wait in each memory pool of the nodes, this transactions stay there until miners comfirm them which is going to be depending on the fee size sat/vbytes you pay for your transaction. Once the transaction is mined and comfirmed by the miners, the confirmed transaction is propagated back to different nodes as a copy to their memory pool.
Each node has history of transactions from the Genesis block to the current comfirm blocks of transactions but some node do prune their node to have more space in their memory because right now, blocsk of transaction from Genesis block is more than 500 gigabytes but there are some nodes that has these blocks of transactions in their memory pool. This has answered your last question.
To your first question again!
What you need to know is mempool.space(explorer)
≠ general mempool of transactions, each Bitcoin node has their own mempool(memory space for transactions for both comfirmed and unconfirmed). However, mempool.space another Bitcoin node run by someone providing the same service of explorer.
My recommendation, if you have a laptop device with atleast 700gb to a terabytes, download Bitcoin core to see how fun is running a Bitcoin node, you will understand it better in practical form.