The ancestry of Otgiva is certainly worth a conjecture or two, if only
readers bear in mind the reservation you expressed in the first place
and that I also emphasise now.
Mathilde of Burgundy may fit chronologically as a candidate for Otgiva's
mother, but circumstantially this does not seem very likely to me.
Mathilde's daughter Berta can hardly be placed as another Luxemburg
(whether Giselbert or Frederic was the father of Otgiva and her recorded
sister Gisela), as in that case if Berta's only husband was a brother of
Pope Leo IX then the possession of Geneva by her son Gerold would become
even more mysterious than it already is.
Otgiva's sister Gisela appears to have been having children as late as
ca 1040 - her son Radulf was killed in battle in 1102 and her son
Gilbert de Gant was father of Robert who lived until 1157/58.
The marriage of Otgiva to Balduin IV is often placed by 1012 - it has
been placed as early as ca 1005 by some (including Nicolas Huyghebaert)
but this is not based on any evidence that I can find. The 1012 dating
appears to be deduced mainly from Christian Pfister's unwarranted
assertion that Otgiva's son Balduin V was married to Robert II's
daughter Adela in Paris at an assembly held in the early months of 1028,
assuming that he was 15 at that time and so born in 1013. However, there
is better evidence that the wedding of Balduin V and Adela did not take
place until after the death of her father (in July 1031), and better
again that it was before 5 July not later than 1034 (in the lifetime of
his father who died in May 1035). It is said to have been brought on by
Balduin's burgeoning manhood, so that he was presumably not far off 17
years old in the early 1030s and this would place the marriage of his
parents by ca 1012/16, fitting well enough with Otgiva's birth by ca
1000. Balduin IV may have been older than her by 15 years or more.
Apart from the probable English connection in her name, Otgiva may have
had a noble Byzantine ancestry through her unknown mother as reputed for
her son Balduin V according to William of Poitiers. This evidently did
not come through Balduin IV, so that Otgiva was more probably the
conduit if this near-contemporary belief was correct. The comparative
modesty of the claim, of noble rather than imperial blood from
Constantinople, does not have the ring of outright fiction. If true the
most likely explanation seems to me that Giselbert of
Luxemburg-Vaudrevange had married by ca 1000 a lady whose parents had
English and Byzantine antecedents between them. Possibly her father was
a son of King Edward the Elder's daughter (Eadgifu/Otgiva) who had been
sent to Germany with Otto the Great's wife Eadgyth, and her mother may
have come from Constantinople in the entourage of Otto II's wife
Theophanu in 972.