New CBB

Which language has the largest consonant cluster inventory? Or what is a plausible, maximal consonant cluster inventory?

Nobody speaks of "consonant cluster inventories". There are inventories of consonant phonemes; and there are clusters of consonant phonemes. People talk about those, but not about inventories of clusters of phonemes.

That would be !Xoon.

ǃXóõ (or however it's supposed to be spelt) is (one of?) the languages with the biggest consonant inventory, yes.

As for consonant clusters, that depends what you mean.
Bella Coola famously has long grammatical complete sentences with many consonants and no vowels.
If you want "biggest consonant cluster in a word", I don't know, but I bet it's either one of those above or one of those mentioned below.
If you want "biggest consonant cluster in a syllable", it's probably a European language; English and German are both good candidates. One allows 4-consonant onsets and 3-consonant codas; the other allows 3-consonant onsets and 4-consonant clusters. Technically, I suppose, they both allow 7-consonant clusters in a word (but a syllable boundary has to occur within such a long consonant cluster).

Nobody speaks of "consonant cluster inventories". There are inventories of consonant phonemes; and there are clusters of consonant phonemes. People talk about those, but not about inventories of clusters of phonemes.

IMO the longest plausible syllable-onset-cluster and the longest plausible syllable-coda-cluster are both 4 consonants long; and IMO the longest plausible consonant cluster in a word is 7 consonants long. But, I believe some conlangers think 5-consonant onset-clusters and 5-consonant coda-clusters are (maybe only barely?) plausible; and if a 'lang had both, it could conceivably have 10-consonant clusters inside a word.

You should know, though, that 3-or-more-consonant onset-clusters and 3-or-more-consonant coda-clusters both co-occurring in the same language is a distinctly minority phenomenon; there are few such natlangs.

Maybe they meant something like that table?

Georgian has some pretty serious onset clusters but it's distinctly lacking very many word final ones, at least in my experience with the language.

Which one is better?
/x/ ‹x› and /ʃ/ ‹c›
or
/x/ ‹c› and /ʃ/ ‹x›

I am redoing Ngith, and am currently deciding the phonology.

The language is planned to have a rich ejective system, and I've made [glottal stop+stop/fricative] a realization of ejectives when whispering. So /ts'/ is [ʔts] when whispered. However, what's preventing someone from postulating that the actual phonemic unit is /ʔts/ and not /ts'/?

Basically, what can I do ensure that these consonants are considered phonemically ejectives and not [glottal stop+stop/fricative] sequences?

The language doesn't even have a phonemic glottal stop in the first place.

Huh? What's preventing a lang from having ejectives but no glottal stop?

There's minimal pairs that help prove that the ejectives are phonemic, however.