I'd try to work it out but man its taking up too much time even just for 2 simple servers on Amazon.

Its so hit and miss too. For instance: Node 2 didn't have any mining payouts from 7/15 to 7/20 having gone down on 7/19. Couldn't revive the node no matter what I tried so I had to transfer the stake & rebuild the node on another local wallet/PC and another remote server on the 20th. It was up without incident but then nothing happened from the 20th to 27th. However from the 27th to the 1st  it did about 6.9 DRK over 6 days which is great if it was consistent; then it'd actually be profitable and worth the amount of troubleshooting effort involved. As far as luck: they seem to be streaky. It was on a hot streak but as of last night I can't seem to remotely log into the server itself for some reason.

Node 4 also saw that same inconsistency. This was actually the most stable node of the four I had set up, but that didn't help its profitability much IMO. It "only" needed one re-activation during 2 weeks and only proved problematic last night on 8/1 when I couldn't log in to the remote server itself. There were 2 payouts on the 19th, then nothing till the 25th six days later, then nothing again until the 1st last night, meaning there were only 4 payouts for a node that was relatively stable in that it only had to be dealt with twice (the second time "fatally", unfortunately) from the 16th to the 1st.

At the moment my best bet would be to do another 2 server rebuilds since as of last night I can't get into either instance even after rebooting them. Regardless of the alleged safer cold wallet method introduced by RC3, I'm tired of fighting these things at a loss and at risk with everyone's funds.

I work in labs and field tests of mostly low Tech Readiness Level products and prototypes that are usually between TRL levels 2-7. For a consumer products example: a new refrigerator model would likely be TRL 9-10. I'm comfortable working with and troubleshooting or assisting with troubleshooting various types of early beta software and hardware. Some would see it as a bad sign that an ubernerd like myself that makes a living in R&D no longer wants to lose money managing beta masternodes.

Nodes 2 and 4 should be seeing payouts in a few hours. The refunds should arrive in the evening after I get back from the beach.

Final tally:

Node 1: RC2 Hot wallet method; stake funds never sent back from mempool when server shut down/rebuilt prior to RC3. Owned 45% stake in this node and no idea who has control of the address the getaccountaddress 0 command yielded on the local wallet with RC2. Asked for help but ignored by Evan.
Node 2: RC3 (as are Nodes 3/4). Refunding. Luckiest of the nodes but still a net loss to operate even with low cloud hosting fees. 7 rebuilds in July, about 10 rebuilds total. Ended w/ inaccessible remote server which doesn't appear to be MN software related but can't be rules out either.
Node 3: Was operational from about 6/25 to 7/2. Database error appeared on 7/2. Phantom self-payment upon rebuild to inaccessible self-address not listed in receiving addresses. Shows up as grayed out address in parentheses but balance of payments equal 1000DRK instead of 0DRK. Copies of orig. wallet file and database/blockchain weren't helpful as soon as it syncs to 7/2, the phantom self-payment appears rendering the stake as unreachable. No idea where this DRK is or if its recoverable. Community help (including offers to pay for help) didn't yield a fix for something that could happen to anyone that sees a database error.
Node 4: Refunding. Most stable of nodes, needed 2 rebuilds in July and 1 re-activation. Had highest uptime of nodes set up but earned less than node 2 in latter half of July.
Node 5: Refunding, was half full and could've been filled easily with community support but declined to host with problems shown and costs of running nodes taken into consideration.

Full Node setups AND management: Canceled.

Full Node Setups ONLY: Available shortly. After refunds are complete, will shift to only doing cold wallet setups for 1 or 2 full node setups (since multiple wallet setups are still somewhat complicated) where I don't have to touch stake funds. More importantly: I won't have to be responsible for stake funds. Except for one large order: Management will have to be self-management or management by a 3rd party. Client will need to set up a host service and allow me temp. admin access. Setups should take less than an hour providing the hosting service is set up and 1000DRK is ready.

MN Repair services: Considering this for a worthwhile fee but would rather just set up nodes and turn over admin access going forward.

All 4 nodes operated at a net loss from 6/25 to 8/2, with 2 of 4 nodes stake funds currently (and possibly permanently) inaccessible without some sort of developer (or divine) intervention. I wasn't trying to get instantly rich from offering this service but I can't operate at a net loss with what appears to be a coin that has twice showed me whimsical and capricious loss control of hard earned stake funds from external/internal factors at a 50% rate, and requires excessive babysitting for the nodes that do work without major issues.