"Glottality" could be measured in a number of ways. First this could refer to glottal stop as an independent consonant, as exists in Arabic. It could also refer to an accompaniment with other consonants (usually stops), in which case the consonant is usually called an "ejective", as in Georgian, Navaho, Amharic. It can also refer to an accompaniment with a sonorant (liquid, nasal or vowel). Especially in the latter case, "glottalized" vowels may be very hard or impossible to distinguish from glottal stop consonants adjacent to a vowel. As a subcase of vowel glottalization, this can be a feature of tone systems such as Vietnamese nặng and ngã tones or Latvian "broken tone". Danish stød is related, but it is hard to classify / analyze.
A second dimension of glottality is predictability. Glottal stop in some dialects of American English is very predictable, appearing at the beginning of a word that seems to begin with a vowel and maybe retained in phrases like "the apple" (depends on whether you say [ðə]] or [ði]); and in place of /t/ in "rotten". In some dialects, word-final /t/ after a vowel goes all the way to [ʔ], thus [kʰæʔ] "cat". Very many languages have automatic glottal stop before utterance-initial vowels, and its actual presence is highly variable. North Saami is another language with quasi-predictable glottal stop. There is a lexical contrast between plain and laryngeally-marked consonants, where this is realized as preaspiration on stops and glottalization → glottal stop on nasals ([laʔnja, bĭemʔmu]) "room; food". In some Swedish dialects, preaspiration is instead a glottal stop, so [goahti] → [goaʔti] "house". Actual glottal stop as a separate consonant is governed by rules, but as a practical matter, it's not "predictable" in the sense that language-learners have in mind.
In some languages, /ʔ/ is a full-fledged consonant which can't be predicted. It is one in Arabic (which does have a rule of glottal stop insertion, but not all glottal stops come from that rule), and in some modern (Levantine and Egyptian) dialects it comes not only from original /ʔ/ but also from /q/.
Glottalization as an accompaniment of other things is usually not automatic / predictable from something else. Ejective consonants in Salishan, Quechua, Navaho, Georgian simply have to be learned. However, you can learn, in Southern Bantu languages that the non-aspirated stops are ejectives – that is, plain unaspirated /p/ is pronounced [p']. Or, you can decide that the language has no plain unaspirated stops, it only has ejectives – it becomes a matter of theoretical analysis, and there isn't a compelling argument for one view versus the other. Glottalization in Vietnamese nặng tone is just one part of defines the tone.
In your list of languages, Italian would be "least glottal", Cantonese "most glottal", and I would put German and English in the middle. One problem is that these are languages with many dialects ("Italian" really is not one language, it's a cluster of Romance languages spoken in Italy). Some dialects of English have more instances of glottal stop, some have fewer. Levantine Arabic probably has the highest occurrence of undeniably underlying glottal stop as independent segment amongst the "easily accessible languages".