Altcoins will not cease to exist but continue to co-exist with whatever Bitcoin is bringing to the table. Lets not forget that Ethereum came up with concept of smart contracts and DApps and we can never count it out. Other coins also have their distinct functionalities with respect to their own chains.
You do have a point there, mate. Regardless if sidechains get implemented into Bitcoin or not, there's a possibility that altcoins would continue to be used the same way as they are right now. There is a wide array of altcoins, each with their unique technologies and features that won't be able to become implemented into a Bitcoin sidechain. Of course, it could be possible to implement any amount of sidechains for Bitcoin to bring many features of altcoins, but the effort wouldn't be worth it considering that altcoins can already do the same functions.
Despite this, it will be very interesting to see a sidechain get implemented into Bitcoin, as it'll can have the same security as Bitcoin itself, but with less load and reduced costs than the main chain. Rootstock is one example, which aims to bring smart contracts functionality to Bitcoin, making it a strong competitor against Ethereum in the future.
Lightning network is far a head from being implemented,theres a lot of problems lately which delays the development of the LN network,altcoins are going to be worthless as soon as the LN has successfully implmented to bitcoin, because most of the altcoins wont survive what will happen to the bitcoin's tx speed which makes most of the investors invest more to bitcoin because there will be less fees,faster tx confirmations etc.
Probably. I think that once Lightning Network becomes actively used by individuals in the mainstream world, they'll use it more than altcoins themselves. Of course, they'll still co-exist with Bitcoin, but the level of transactions made on Bitcoin would be greater with the Lightning Network than with altcoins. If successful, this could render Bitcoin Cash useless as costs would remain low, and speed would be as fast as possible without putting much stress to the network by upgrading the block size. Just my thoughts