When an upgrade from 14.04 LTS is convenient or necessary?

I am on a perfectly configured and functioning Ubuntu 14.04 LTS machine.
LTS versions of Ubuntu are supported for 5 years, so in theory there should be no need to upgrade the whole system before the 5 year deadline. I conceive the whole point of having a LTS system the best choice for those like me who prefer to rely on stability for as long as possible, because they develop software or web application for instance: a system upgrade is always a tricky operation.

Considering that I am aware that when 2019 comes Ubuntu 14.04 will not be supported any more, and that I certainly will want at that time to extend the life of my system, I still do not understand what possibilities and options am I left with.

Will in 2019 I be able to upgrade from 14.04 to 18.04 or it is necessary to upgrade gradually, when time comes, from a LTS to the next one? In other words is it necessary now, or in about three months time, to upgrade from 14.04 to 16.04.1 in order to be able to upgrade in 2018 to 18.04?

What is the best and safest choice I can do?

I don't know if you'll be able to skip an LTS. I think it most likely you'll need to hop via 16.04. But you can do both those steps in 2019. You don't have to move to 16.04 now. You can, but you don't have to.

There may be an upgrade path plan published for 14.04→18.04 so that might be an option, but it doesn't make any difference in practice. Your job is the same. You're going to change the base and you'll need to test your software on it.

The important lesson to consider here is you can test whenever you like. Leaving that until 14.04s is dying, and finding out you need to do weeks of improvements is probably a bad idea. Doing an untested upgrade on a production system and then finding out it's broken is much worse.

There is nothing stopping you, for example, creating a testing system running whatever version of Ubuntu, that you test on. That will let you know what doesn't work and you can incorporate that into bugfixes, or other plans for when you actually deploy. You can go one further and use "continuous integration" techniques to find out where the problems are, as they happen. I'm not sure how appropriate this is for you though.