Convert To Layers - keep size and position

One work-around to convert one instance of a Linked Smart Object into an Embedded Smart Object (edit: which could be automated via a Script): 

• Image > Duplicate 

• Layer > Smart Objects > Embed Linked 

• Layer > Duplicate (to the original image) 

• close the duplicate without saving

• delete the original instance of the linked SO

I agree, this can be frustrating. A workaround is to use ‘Rasterize Layer’ instead of ‘Convert to Layers’ to maintain the transformed size. Alternatively, you can create a smart action that duplicates and rasterizes the SO before converting. You might also find discussions on similar issues helpful here: [Adobe Community Link, if available]. Hope this helps!

The features work as they are intended to work.

But if necessary one can work around that as @Stephen Marsh already explained.

Have you considered using layer comps in the smart objects and applying those in the containing document via the properties panel instead of rasterizing?

Please provide a sample file.

OK so if you have an SO and it is duplicated through several artboards, and you use Relink to File, does it not change all SOs? And if you then try using Replace Contents, do not all SO copies do the same? Because that has been my experience. If I want to change just one of those SOs I copied, that's impossible.

By @Todd_Morgan

We have had this discussion before. That is how smart objects behave. I do understand that you want linked smart objects to behave differently. That would be a feature request "idea" as it's not a bug, it's long-established designed behaviour, originally set with embedded smart objects.

One needs to use the "New Smart Object Via Copy" command to isolate one particular embedded smart object so that it isn't synchronised.

There isn't such a native feature for linked smart objects; as they are linked, all links get replaced. In the previous discussion, I offered the manual steps, an action and a couple of different scripts to automate the creation of a "New Linked Smart Object via Copy" which should solve your problem.

OK so if you have an SO and it is duplicated through several artboards, and you use Relink to File, does it not change all SOs? And if you then try using Replace Contents, do not all SO copies do the same? Because that has been my experience. If I want to change just one of those SOs I copied, that's impossible.

Created a plethora of digital banners in various sizing, 20-30, and in those banners are vrious layouts and scaled usage of SO within them. So if I have in one banner this same SO from another bigger banner, but then need to change something for that one banner having the SO unpack at that new scaled sizing, with all layer effects etc as in situ, that would be cool. Because we all know, the relink to file or replace contents doesn't work to how I would like it, and this is the only option. Or... would love to be able to open into that SO and tweak it, but not have the tweak read by other SO copies in the PSD. I would love for Replace Contents to work where the replacement doesn't change all other instances - Relink to File and Replace Contents do the same thing so waste of right-click space if you ask me

Also: How complex are the transformations, Filters etc. applied to the Smart Object instances in the containing document likely to be? 

»Unpacking« a Smart Object could lead to undesirable results, for example if a Filter is applied to Layer Masks individually instead of on the result. 

To illustrate: 

Me and, I think, several others tried to create Scripting-solutions to »unpack« Smart Objects (though this kind of defeats the purpose of Smart Objects themselves, in my opinion). 

But the issue is quite convoluted – transformations, warps, puppet warps, Filters etc. applied in the containing document; Layer structure, Vector Masks, Mask settings in the Properties Panel, blend settings etc. in the Smart Object; and to top it off: differing Color Modes and bit depths that may make the transfer of certain Adjustment Layers truly impossible. 

What is the position this functionality would play in your workflow? 

Why do you need to disassemble a Smart Object into its constituent parts within the containing document? 

How complex are the Smart Objects you are trying to dismantle (on average at least)? 

Could you please post screenshots with the pertinent Panels (Toolbar, Layers, Options Bar, …) visible to illustrate?