EDIT

Errors in Data:

• Name inheritance:

Some (63) children differ in first name part from parent. Sometimes this is simply a wrong whitepace, but more often it is a misspelling:

   SELECT syn.id, synParent.tu_name parentEpi, synParent.tu_displayname parentDisplay, trim(replace(replace(replace(left (syn.tu_displayname, len(syn.tu_displayname)- len(syn.tu_name)), ' var.', ''), ' f.', ''), ' subsp.', '')) shorted,
       syn.tu_name synName, syn.tu_rank synRank, syn.tu_displayname synDisplay, syn.tu_unacceptreason, right(syn.tu_sp, 30) synTree, syn.tu_parent synParent,    syn.tu_status synStatus
   FROM tu syn 
     INNER JOIN tu synParent ON synParent.id = syn.tu_parent
   WHERE syn.tu_rank >= 220 AND synParent.tu_displayname <> trim(replace(replace(replace(left (syn.tu_displayname, len(syn.tu_displayname)- len(syn.tu_name)), ' var.', ''), ' f.', ''), ' subsp.', ''))

Subgenus seems not to have this problem and for genus and above it is not relevant anyway:

SELECT syn.id, synParent.tu_name parentEpi, synParent.tu_displayname parentDisplay, trim(left (replace(replace(syn.tu_displayname, ' (', ''), ')', ''), len(syn.tu_displayname)- len(syn.tu_name)-3)) shorted,
       syn.tu_name childName, syn.tu_rank childRank, syn.tu_displayname childDisplay, syn.tu_unacceptreason, right(syn.tu_sp, 30) synTree, syn.tu_parent childParent, syn.tu_status childStatus
FROM tu syn 
  INNER JOIN tu synParent ON synParent.id = syn.tu_parent
WHERE syn.tu_rank <220 AND syn.tu_rank >180 AND synParent.tu_displayname <> trim(left (replace(replace(syn.tu_displayname, ' (', ''), ')', ''), len(syn.tu_displayname)- len(syn.tu_name)-3))

• there is one case, where a nomen nudum is the accepted taxon for an other taxon

SELECT  acc.tu_status, acc.tu_displayname acc, syn.tu_acctaxon, syn.id , syn.tu_displayname syn
FROM    tu acc INNER JOIN
               tu AS syn ON acc.id = syn.tu_acctaxon
WHERE     (acc.tu_status = 3)  AND syn.tu_acctaxon <> syn.id 

• there are 3317 cases where the tu_acctaxon has not a status "accepted" (2213 are self-referencing, 1104 are really different records, none of the "accepted" link again to another record, though there seem to be such records in WORMS, e.g. http://marinespecies.org/aphia.php?p=taxdetails&id=210439)

SELECT  acc.id,   acc.tu_status AS accStatus, acc.tu_displayname AS accName, 
            syn.id, syn.tu_status AS synStatus, syn.tu_displayname AS synName
FROM    tu acc INNER JOIN
            tu AS syn ON acc.id = syn.tu_acctaxon
WHERE     (acc.tu_status <> 1) -- AND  syn.id = acc.id --  AND acc.id <> acc.tu_acctaxon
ORDER BY accStatus, synStatus 

• 289 cases where parent taxon is not accepted => probably not critical assuming that the parent child relationship is not a taxonomical one

• 610 taxa being alternative names but accepted taxon has not status accepted. Especially critical for both having status 5 (alternative name)

SELECT  acc.id, acc.tu_status AS accStatus, acc.tu_displayname AS accName, 
            syn.id, syn.tu_status AS synStatus, syn.tu_displayname AS synName
FROM    tu acc INNER JOIN
            tu AS syn ON acc.id = syn.tu_acctaxon
WHERE     (acc.tu_status <> 1) AND (syn.tu_status IN (1,5,9,10))  AND  syn.id <> acc.id -- AND acc.id <> acc.tu_acctaxon
ORDER BY accStatus, synStatus 

699 accepted taxa (acctaxon = id) do have a parent which is not accepted. 2/3 of the parents has status synonym, 1/3 has status alternate representation.
This is CRITICAL as these children are difficult to handle. => not really true anymore with the current strategy to build tree only from tu_accfinal accepted taxa

SELECT  child.tu_status childStatus, child.tu_displayname child, substring(child.tu_sp, 2,1) kingdom, child.tu_rank childRank, child.id childId, child.tu_acctaxon childAccId,
 par.tu_status parStatus, par.id, par.tu_name, par.tu_displayname, par.tu_acctaxon, par.tu_rank, accPar.tu_displayname accParent
FROM  tu child INNER JOIN tu par ON par.id = child.tu_parent INNER JOIN tu accPar ON par.tu_acctaxon = accPar.id
WHERE child.tu_acctaxon = child.id AND par.tu_acctaxon <> par.id -- AND par.tu_status = 2
ORDER BY  kingdom, child.tu_rank, par.tu_status

There are 82 tu_status=accepted taxa with unaccept reason: (mail sent)

SELECT *
FROM tu
WHERE tu_status = 1 AND tu_unacceptreason IS NOT NULL

ACCEPTED TAXON:

336 taxa have NO accepted taxon (tu_acctaxon IS NULL)

57003 are their own accepted taxon (tu_acctaxon = id) (tu_status : 1=54790, 2=8, 3=10, 5=29, 6=410, 7=71, 8=385, 10=1300)

35706 have another accepted taxon (tu_acctaxon = id) (tu_status: 2=33804, 3=39, 5=1842, 6=21)

SELECT *  --  tu_status, count(*) as n
FROM tu
WHERE tu_acctaxon =/<> id/IS NULL
-- GROUP BY tu_status
ORDER BY tu_status

NO synonyms exist with an accepted taxon again being a synonym of another accepted taxon:

SELECT  acc.id accId, acc.tu_rank, acc.tu_displayname accName ,acc.tu_sp, acc.tu_parent, acc.tu_status,syn.id, syn.tu_name synName, syn.tu_rank
        synRank, syn.tu_displayname synDisplay, syn.tu_unacceptreason, syn.tu_sp synTree, syn.tu_parent, synParent, syn.tu_status synStatus
FROM tu syn LEFT JOIN tu acc ON acc.id = syn.tu_acctaxon -- AND syn.id <> syn.tu_acctaxon 
WHERE  
 syn.id <> syn.tu_acctaxon 
 AND acc.id <> acc.tu_acctaxon