If you extract fairly clean pseudo HCl, you may freebase it with sodium carbonate and extract into any suitable nonpolar solvent. There is no need to use the salt, or for that matter the NaOH, or the carbon. The NaOH is overkill; the carbon and salt are no longer needed if the pseudo has already been extracted by a good method. To extract the freebase use naptha and freeze to precipitate or use tolulene and evaporate.
You may use water to mobilize the base if you are sure the pseudo is clean of PEG and Orange gakks. If it is not, using water to base will most likely not solve the problem.
Any good waterless a/b technique will work here, and when doing this I prefer to mix dried alcohol and dried acetone to mobilize the base to freebase the pseudo HCl. Note that if you are already orangeII gakked, the pseudo may not freebase with sodium carbonate, mobilized with the alcohol/acetone fluid or with water. I have observed this phenomenon twice now. On both occasions when sodium carbonate did not work to freebase the pseudo, orange II contamination was present. Just an observation. I have no idea about causal relationship.
I suggest that prior to freebasing the pseudo HCl you determine how clean it really is. Burn tests help. You may also note how fine and powdery the pseudo is. Clean pseudo will crumble and pour, break apart without resistance, dissolve readily. Dirty pseudo tends to clump, cling to the sides of dishes, dissolve slowly, and not crumble. If you have dirty pseudo, IPA/acetone recrystalizing helps clean it, as does the xylene clean. I have discovered that the xylene precipitation technique post Straight to E extraction will clean most remaining gakks from the pseudo HCl. whether this is due to fresh solvents or the absence of other gakks the second time around I do not know. I just know it works well.
Cleanliness of the pseudo will help improve yields. I have discovered that my yields from pseudo HCl and pseudo freebase have been equivalent in the PIE long wet reflux when the pseudo was equally clean. The real advantage to reacting the freebase form may stem as much from the better cleanliness of the freebase form as any thing else. After a couple of recrystalizations after the STE process, yields have been equivalent to yields from freebase. Yields from some pills after STE alone have been better than my best yields from freebase. I duplicated this with pseudo HCl from the same pills; I have not had the same results with any other pills to this point. This suggests that the cleanliness of the pseudo, whether HCL or freebase, is likely more important to yield that the form of the molecule. Freebase pseudo as a rule tends to be cleaner than most pseudo HCl extracted from pills, and that may explain why many of us report higher yields with freebase.