The best practices for notating polymeters are highly contingent on the specifics of what you are doing, so...
After discussion, we've established the following constraints:
- there is an ostinato part of 3 + 3 + 4 + 3 semiquavers (sixteenths);
- the main parts will be working polymetrically against this ostinato;
- there will be parts that sporadically reinforce (but don't fully double) the ostinato.
Because it is a multi-instrumental score, you will need bar lines that coordinate all the parts, so, yes, everything will be forced into one set of bar lines. You will thus need to use other cues to establish the rhythmic relationships between the parts. Using What do Four Vertical Dots mean? as a guide, and adapting Bartók's methods to suit what you are doing, I'd suggest:
- beaming the ostinato by the groups of 3 and 4 semiquavers, beaming across the bar lines as necessary - this should make the ostinato's rhythm very obvious;
- adding tick or dotted bar lines to all the accompaniment parts that follow the ostinato's rhythm.
You don't need to be as abstract in your coordinating meter as Bartók was. I'd use the main parts to establish the overall meter.
I suggested perhaps adding colla parte referencing the part that states the complete ostinato to the parts that need to follow it. Colla parte ("glue the part") is usually used for the accompaniment of a main part that is playing molto rubato (i.e., follow the main part's timing), but is perhaps applicable here. However, on reflection, I would just suggest adding the complete ostinato as a cue line to the parts that rely on it when you come to extracting the parts.