Whatever colonists you send there will get regularly smashed by the same hordes that currently smash PeL. Except that their income will be !@#$ty, and won't have any population to support decent recruitment rates.
Not true. They will be smashed by those same hordes minus whatever hordes may have spawned in their territory.
Moreover, if they hold on for long enough, the spillover could result in PeL being able to take more regions to hold, further reducing spawn rates. The exact size of the network effect would be disputable.
However, a much better location for such a "spawn reduction realm" would be Sallowtown. More easily defensible, probably better income (though still pitiful), and a more direct route to a "border close." The new realm could TO over to Axewild, while PeL TOs to the coast. If Luria Nova simultaneously spawned a new realm in Shinnen, big things could happen. The northern border of "Greater Luria" would be Desert of Silhouettes, Axewild, Lupa Lapu, and Poryatu or Dantooine. A capital shift to Poryatown would further ease defense, as would a new realm in Shinnen. This would establish "corner capitals" at Sallowtown, Poryatown, Shinnen, and Giask. Forland being essentially pacified between the Shinnen realm, Luria Nova, and Fissoa, resources could be committed northwards to allies in Sallowtown or Poryatown. Under these conditions, Flowrestown or Balance's Retreat could be meaningfully colonized.
The trick to beating the monsters is maximizing "secure regions" (regions with no rogue borders) and minimizing "border regions," minimizing travel times, and TOing in patterns that achieve the aforementioned maximization and minimization. Moving to R/CTO Balance's Retreat does not reduce the number of human border regions or increase the number of human secure regions, and creates big travel times for support. Moving out on the flank to Sallowtown could accomplish this (as well as being more easily reinforceable by potential friendlies from D'Hara).