Played around a bit with it last night, and overall the edit seems to be on the right track. Here's a few errant thoughts n' improvement suggestions based on ~15mins' worth of playtesting.
- Vegetation collision was the most ubiquitous remaining issue I ran across, especially while on the hoverboard: plenty of shorter plants where one has to wonder what the gameplay point of enabling their karma/actor blocking properties was. Beyond this, the taller trees' downwards growing branches/foliage also collide with actors, pedestrian or otherwise, unexpectedly obstructing movement through 'em and allowing for some ...unconventional, acrobatic tactics should one be inclined to start shield jumping up them high enough. I'd recommend a collision disabling spree across the islands' flora with a mind of improving overall gameflow.
- Reworking the bases' geometry was long overdue and this concept seems serviceable enough, if somewhat lacking in z-axis tactical variety. Perhaps consider moving the antennae scaffolds towards the center of the map a bit in order to create enough space for a square hole above each core to be made as an alternative means of attacking, plus some (obvious to defenders) means of getting up on the core structures' roofs? Additionally, maybe place a couple of new rock st.meshes nearby the existing, bigger ones directly opposite each core in order for defenders to be able to climb up there and counter enemy core attackers?
- Possibly flatten the beachfront nodes (i.e. zero their pitch n' roll values) and do some more playtesting to ensure no obstructed angles remain from where players are unable to walk on the node bases.
- Link ammo seems a bit scarce at 100 per locker, kinda undermining the point of the CustomWeaponLockers' inclusion. Bump up to, say, 170ish?
- With only a dmg. amp on offer, the map comes across as pretty stingy in terms of helpful goodies, rewarding exploration and/or helping break through [bigger crowd] defenses. Yes, I'm aware of the few kegs n' super shields, and not inclined to recognize them as particularly useful in the generally uphill struggle towards the enemy core. Maybe judiciously sprinkle a couple more tactically useful items here or there? Possibly add a pair of symmetrical, far off islets for the hoverboards to go n' grab an ion or a translocator or such; or, alternatively, follow the ONS-Tropic-32p example and create a patch of beach/land right behind each core's rocky formation so something useful can rest there? Oh, and btw, I forgot to go over the mine "economy", so someone plz check that it makes sense rather than being too stingy or overly generous n' lag-inducing (remember, mines can be placed atop hoverboards too).
- On a lark, maybe add a few more location names and node descriptions? Currently, the locational information accuracy exhausts itself at the "east/west" point.
- Lastly, I must admit I've yet to become a proponent of the EONS Locust. It served an interesting purpose in UT3, what with its accompanying grapple functionality, but, divorced of such a mechanic here, the UT2004 port/equivalent manages to offer... what exactly? A less nimble, more vulnerable, non-strafing take on the manta, from which its driver can be separated after receiving any hit? Why not explore whether mantas can reasonably [or additionally] be accommodated in SavoIsland? Ditto for a bender or scorp per beach primary. Nothing crazy, but I think the map should be able to handle a total of 4-6 stock, lower-tier vehicles well enough.
That's all I got, hope it helps. Props for current progress and keep up the good work to all involved editors n' content contributors

.
Eyes in the skies.