One of the solution to address low latency and high scalability in BTC is to increase block size. A single block takes 10 min on average for proof-of-work. Increased block will also take same time for proof of work and network latency will be increased. A transaction when recorded in Blockchain will become valid only after proof of work is validated through consensus. I fail to understand how will increased block size decrease time for transaction to become valid and recorded in ledger. At max, It will just allow more number of transactions in any given block.
Block time and block size are two separate things. Both of them have their own effect on scalability (if you're considering transactions per second as the key factor of scalability)
If you zoom out a little and look at the scenarios from a 60-minute point of view, a larger block size (let's say we go from 1 MB to 2 MB) allows twice the number of transactions to go through in the same amount of time. The block time for both 1 MB and 2 MB blocks remains 10 minutes, but 2 MB blocks increase the network's transactions per second (or per block, if you want to count that way) to twice the previous amount.
If in this network you also halved the block time to 5 minutes, you've effectively quadrupled the throughput of the network.
Increasing block size is not meant to decrease the minimum time required for a tx to be recorded in the blockchain. It is meant to decrease the average time taken. Since bitcoin currently has full blocks, increasing the block size means that more transactions are confirmed in the same time, thus decreasing the average amount of time it takes for a tx to be confirmed.
Transactions are valid before they are added to the blockchain. However, they only become reliable by being confirmed in a block that ends up being part of the heaviest chain.
Even if blocks were to have more capacity, the (optimal) waiting time until first confirmation would not be changed by that. At best, more transactions would be confirmed in the next block, which may reduce the average waiting time for first confirmation. Since most transactions are known to the network before they are confirmed in a block, and compact block representations have been added to the protocol, the latency of block announcements does no longer increase linearly with bigger blocks.
To achieve faster first confirmations, the block interval would need to be reduced, which at this point in time would seem to be a difficult change to get accepted by the network.