Let's talk about the cause of your issue first.
The reason you see your 'in-stock' computers automatically changed their locations back to their original locations is that the scheduled task Network Resource Location Update kicked in and it changed your in-stock computers' location back to the locations associated with their IP addresses (subnet).
So, you can either stop this scheduled task from running...which you will certainly give me a .
Alternatively you can exclude these in-stock computers from being flipped. I have created a video to demostrate how you can achieve this in 7.x environment. I will post it here once it is available. However, you are on Asset 6.5, but the theory is exactly the same.
The solution is to create a collection (in 6.5, filter in 7.x) with the following sql query to include all in-stock network resources:
declare @inStock uniqueidentifier
set @inStock = '1C139F6C-F210-4002-90D0-4DFAF98D5FA4'
declare @assetStatus uniqueidentifier
set @assetStatus = '3028166F-C0D6-41D8-9CB7-F64852E0FD01'
(For 6.5 users)
select ri.Guid from vResourceItem ri join ResourceTypeHierarchy rth on rth.ResourceTypeGuid = ri.ResourceTypeGuid join ResourceAssociation ra on ri.Guid = ra.ParentResourceGuid and ra.ResourceAssociationTypeGuid = @assetStatus
where rth.BaseResourceTypeGuid = 'ECADDEC9-3178-4FFB-BEB4-D29E0767A79E' and ra.ChildResourceGuid = @inStock
(For 7.x users)
select nr.Guid from vRM_Network_Resource nr
join ResourceAssociation ra on nr.Guid = ra.ParentResourceGuid and ra.ResourceAssociationTypeGuid = @assetStatus
where nr.Deleted = 0 and ra.ChildResourceGuid = @inStock