It's not really creating life from scratch, it's copy-and-paste. Being cross-species is kind of interesting, but we've done the same kind of thing before.
Really? Where did we do the same kind of thing before?
People are heralding it as a huge scientific achievement because it is. At the end of the day, it doesn't really matter if we made the entire cell from scratch or just bootstrapped it from an existing cell. The best analogy here seems to be one from computer science -- say you just designed a new C compiler. Now, before you can actually get it to compile things, like
itself you have to...compile it. It obviously can't compile itself without already being compiled. This is known as the
bootstrapping problem.
There are two ways to deal with the bootstrapping problem: you can create a tiny compiler in assembly to compile yours, and then compile your own with your compiled compiler again to complete the process. Or, you can take an off-the-shelf compiler, then compile yours with that, then finally compile your own again with your compiled compiler to complete the process.
This research did the latter method. And it's pretty obvious to see why if you think about it -- if you're doing genetic engineering, it doesn't matter if you made the whole cell from scratch or you just used an existing cell to bootstrap your genome. The end result is the same, and the end result is what we care about, so why on earth wouldn't you take the easier road? The interesting part is the genome you just inserted (the compiler you just made) and the fact that you now have new cells made from that genome (have a compiled version), not how you got it there (compiled it).