Similar to AutoCAD which differentiates between BLOCK which writes to the internal Block Table (Mstn Shared Cells) and WBLOCK that writes to an external file.
Mstn has slowly incorporated Shared Cells but not really reflected this in the Cell Creation workflow, resulting in huge confusion.
See previous discussion: Some Proposals item 1
Short response... The tool already exists as a checkbox in the cell library dialog and is employed at the time the cell is placed (copied) from the library into the .dgn file. I don't see a key-in to toggle it off and on.
Longer response...
Your link didn't lead me to any previous discussion, so I didn't get the background history.
You might consider putting aside the AutoDesk way for a moment and explore the MicroStation workflows because everything you want to do is likely already in place, but not as you expect to find and use it.
The wblock out concept as I understand it, is done in MicroStation with a fence and a key-in, either sf= or ff=. Alternatively, the commands are fence separate or fence file, depending on whether you want the outgoing graphics erased from the active file as the graphics export to a new .dgn file.
The outputted file was never intended to be a cell nor a cell library. This is a concept difference with AutoDesk products, and consequently, an inherent workflow difference. Bentley stores cells in libraries for later use. There are at least 4 types of cells. Autodesk does not compact individual blocks into a library file, they are maintained in individual external files, waiting to get inserted. When you want to add cells into a library, just attach the library and use your graphics with a fence and have to store cells, they store externally in a .cel file. When you place (insert) cells in a .dgn file, they get copied from the library into the active file.
These are different approaches to storage, requiring different workflows and tools, but accomplishing the same purposes.
Also, cell placement is either as a [static] cell or a [space saving] shared cell (definition), which has its corresponding instance(s) of place, i.e., scale, rotation & location.
One just makes a cell, then places the cell into a .dgn file either with [X] Use Shared Cells toggle turned ON or OFF.
You either get a regular cell or get a shared cell definition placed into the .dgn file, plus the shared cell instance, which is the visible graphics part. Subsequent placement of the same shared cell only adds another instance, and the original shared cell definition is kept to reference. Cells can be replaced to pick up a new definition. Replacing the shared cell definition will change all the instances, that's what the sharing is all about, plus for saving space. You can do BOTH methods in the same file simultaneously, static + dynamic/space saving!
Cells were never meant to be Autodesk compliant, so there are different procedures that need to be discovered.