I would like to raise a little concern here, I do not think it is necessarily a good strategy to pick two very broad topics or terms and decide about a beginner's project by simply looking at what is 'hot'. Without further background it is impossible though to give much more advice about a specific project for you, you must understand that metagenomics and genome editing are still very broad topics and that a beginner still can have a varied skill set.

That said, I would personally opt for a metagenomics topic, simply because it requires more classical bioinformatics, and the bioinformatics pipelines for metagenomics are quite well defined. What is the bioinformatics content of genome editing? It is mainly a set of lab techniques in my understanding with some requirement for designing primers, maybe predicting good targets. For selecting a first step project it is more important IMO that it is well-defined, self-contained and comprehensible, based on existing software and literature. A small metagenomics data set (bacteria, not viral) might be like that.

[Should someone] start with these?

I have no idea what you mean by "start with these". If you mean learn about them and/or work on a project involving them then sure.

Do they matter?

Yes, just search pubmed.

Are they hot?

Who cares, but genome editing is particularly hot at the moment.

Are they important?

The better question is if you find them interesting. If not, you should work on something else.

What problems can they solve?

Search pubmed, you'll find plenty of relevant uses.

Yes, I would definitely agree with that, however before start searching something serious like this you should look for some variants of topics or see how to do it at chemistry research paper topics . Only after the topic is chosen, you may start looking for the information and the sources.