Dear Harry,
Since I have not yet seen a reply to this, and although I am not a Vedic specialist, I will say that these are not misprints, but are in fact sandhi as you suspected. The final dental "n" on sarv
ān would change to the palatal "
ñ" before the initial palatal "j" of jambhayan. This is a sandhi rule and is the norm in Sanskrit manuscripts. However, an editor of printed texts is free to override the sandhi rule and restore the dental "n" if he so wishes. This is sometimes done for clarity, and is often done when the words are separated by a space in printed texts, unlike in manuscripts. So both are correct.
The added "t" after jambhayan before the initial "s" of sarv
āś is due to an optional sandhi rule in Vedic Sanskrit. On this you may consult Arthur Macdonell's
A Vedic Grammar for Students, p. 30, paragraph 36.a., or his full
Vedic Grammar, p. 69, paragraph 77.2.f.
The nearest thing to a definitive edition of the Taittirīya Sa
ṃhitā is the edition by
N. S. Sontakke and (mostly) T. N. Dharmadhikari, 1970-2010, five volumes in nine parts, mentioned by Peter Wyzlic. It supersedes the three older editions you quoted from (of course, barring misprints). For this phrase it has (vol. 3, part 1, 1990, p. 389): sarv
ān jambhayantsarv
āśca.
Your surmise is correct that Mahadeva Sastri and Rangacharya in their edition with Bha
ṭṭa Bh
āskara Mi
śra's commentary never edited book 4. This is because no manuscript of book 4 with his commentary could then be found. Even for this volume of Dharmadhikari's edition, published in 1990, no manuscript of his commentary on book 4 could yet be found. However, his commentary on book 4, chapter 5, the Rudr
ādhy
āya, was available separately, so is included in this edition. It is much more extensive than S
āya
ṇa's commentary on this chapter.