How do you judge the potential of a newer journal?

To expand upon part of Dave Everitt's answer,

3. look at the founders of the journal - did they start it primarily as a platform for their own publications, and do these authors and their associates make up the bulk of the authors?

I would also look at who is publishing and editing the journal.

Is the publisher affiliated with a scientific society? Sometimes scientific society will launch new journals (e.g., the Ecological Society of America launched Ecoshpere to be their open source journal).

Or another is a new journal from an established publisher? Established commercial publishing companies are launching new journals. Although people argue about the ethics of publishing with them (cf What are some examples of negative effects on a career for boycotting Elsevier journals?), my own experience has been that their products are usually scientifically sound (presumably because they are afraid of hurting their brand image). Additionally, journals such as Nature now have spinoffs (and spinoffs of spinoffs, e.g., Nature to Nature Methods to Scientific Reports) to catch paper that do not fit into their flagship journals.

Also, who is the editor-in-chief and other editors? Do you know them in your sub-field? For example, when the journal Theoretical Ecology launched, the they had Alan Hastings as the EIC, a highly respected Ecologist.

Is your niche field organizing it?

If you are in a fairly niche field, then you should know most of the significant people in that field - and preferably actually know them, not just their names.

I strongly believe that a new niche effort (no matter if it's a journal, conference or association) can be successful and worthwhile only if it's done by those people (or a significant part of them) in order to fill their own niche needs. No matter if they come from different organizations (of course they do) or they haven't a single specific association (which most new fields don't have), there is a "virtual community" of people interested in that field and working on it.

Are the key people you know involved in that journal?

Regardless of that, maybe ask them directly about that journal? In the end, it's their opinion that will determine whether that journal will be successful, not any arbitrary metrics.

A suggestion checklist for new journals

1. check who has been published in the journals - are they notable contributors to the field, or all from one research group?
2. circulation - try to find out how many copies they print (if printed) or readership numbers
3. look at the founders of the journal - did they start it primarily as a platform for their own publications, and do these authors and their associates make up the bulk of the authors?

Having said that, although most fields have their primary journals, and new ones have to start somewhere, you can still use the above to litmus-test a new journal (as well as the vanity journal tests e.g. would they accept your paper without any peer review?).

Finally, beware of the new swathe of bogus journals, promoted by well-disguised spam targeted at academics via their university email addresses, for example, see: Journal accepts bogus paper requesting removal from mailing list(!) and Murky world of 'science' journals a new frontier for climate deniers