Optimal insert size for de novo assembly from PE150 reads

[anth]

In the past, I have had reasonable success with de novo assembly of non-model eukaryotes with insert sizes of ~300 bp, using 100 bp PE and Trinity for assembly.

With 150 bp PE reads and the current version of Trinity, is a mean insert size of ~300 bp still generally optimal, or would a longer mean insert size likely benefit the assembly?

(RNA will be poly-a selected, with libraries constructed using the NEBNext® Ultra™ Directional kit.)

Thanks in advance!