Recently Microsoft has published the results of their one year long research effort to determine what method of scaling is superior in term of future-proof “millions of tx per second” requirement.
And their view of increasing block size, Bitcoin Cash approach to the problem, is negative. Citing the Microsoft's post, such approach is “degrading decentralization” and wouldn't allow further scaling. Director of Program Management at Microsoft’s Identity Division Alex Simons writes in his report:
“While some blockchain communities have increased on-chain transaction capacity (e.g. blocksize increases), this approach generally degrades the decentralized state of the network and cannot reach the millions of transactions per second the system would generate at world-scale.”
He says, that the only viable solution is implementation of “layer 2” protocols. In layman's terms, layer 2 protocol is centered around multiple sidechains being brought under the main chain to further distribute necessary computations, and only the relevant finished result being uploaded to the main chain. If all this sounds to you awfully similar to Lightning Network protocol, then you are right. Lightning Network is, in fact, the proposed solution to the problem, and Microsoft even announces, that they will collaborate in development of layer 2:
“To overcome these technical barriers, we are collaborating on decentralized Layer 2 protocols that run atop these public blockchains to achieve global scale, while preserving the attributes of a world class DID (decentralized identity) system.”
Considering the great success of Lightning Network, which has already surpassed Bitcoin Cash in node count across the globe, it seems, that Microsoft is on to something really benificial. If all the promised qualities of Lightning are true, soon we'll see tx being completed in seconds, and fees in range of cents.