I noticed that the guide did not explain that
writing materials determine the
appearance of a script. For example: cuneiform scripts impressed by reed styluses onto clay tablets, runiform scripts were carved onto stone or wood, and sinoform scripts were painted with a ink brush. Each of these methods had limitations that resulted in their characteristic aesthetic, like a brush painting lines of in/decreasing thickness or wood splitting if carved along the grain or leaves tearing if you try to draw angles. Thus, writing your constructed script with a modern pencil or pen on paper has a very different result than writing with a different implement and surface.
On a related note, the guide neglects to mention
the importance of orthography.
Modern
linguistics classifies writing systems into at least four or five types (consonantal, alphabetic, alphasyllabic, syllabic and logographic). (The ever-schizophrenic Wikipedia uses it own made-up words for these but what I just gave are the most common and more importantly unambiguous jargon in linguistic studies produced in the Anglosphere.) As the names imply, a consonantal script writes only consonants, an alphabetic script writes all phonemes, an alphasyllabic script writes combined consonants and vowels, a syllabic script writes syllables or parts of syllables, and a logographic script indicates a combination of phonemes and morphemes. (Despite the name, an alphasyllabic script is not a combination of alphabetic script and syllabic script but a sister system to the alphabetic script, as both alphasyllabic script and alphabetic script are descended from consonantal script.) Of course, these classifications are ideals and in practice most scripts tend to deviate from this ideal unless regularly reformed by the relevant authority. (The ever-schizophrenic Wikipedia cannot distinguish between different types of scripts that don't fit into its arbitrary ideals.)
Chinese, Cretan, Egyptian, Mayan, and Sumerian are among the oldest independently developed scripts and all of them are logographic (Chinese), mixed logographic–syllabic (Cretan, Mayan, Sumerian) or mixed logographic–consonantal (Egyptian). Studies consistently suggest that human brains naturally interpret speech as syllables rather than individual phonemes (one study found that a syllabic script consisting of separate graphemes for body and rhyme was easiest for the test subjects to learn). Egyptian was mixed logographic-consonantal because it inflected poly-consonant roots by changing the vowels, and from this script almost all modern consonantal script, alphabetic script and alphasyllabic script descend.
Some types of scripts are better suited for certain grammar and phonotactics than others. A consonantal script is best suited to a language that inflects by transfixing, so that even if the inflection is not noted you can figure it out from the surrounding context. Syllabic scripts have difficulty representing languages with complex phonotactics or polysynthetic grammar; most languages still using syllabic script are isolating/analytic or agglutinative and have very simple syllable structure. Linear B, a descendant of Cretan, was used to write Ancient Greek and encountered difficulties representing the complex syllable structure of that language. The Cherokee syllabic script encounters difficulties representing Cherokee's polysynthetic grammar, as the syllabograms obscure the roots of words. Conventional alphabetic and alphasyllabic script are less than ideal for languages with lots of consonant allophones like Japanese (as seen by competition between Kunrei-shiki rōmaji and Hebon-shiki rōmaji), initial consonant mutations like Celtic languages (which makes looking up unknown words in dictionaries nigh-impossible), or have far fewer syllables than their phonology/phonotactics allows as seen in Chinese (compare conventional Pinyin to
a purely phonemic model).
An "onset-rime" syllabic script divides syllables by onset and rhyme, which I did not mention above since it was seemingly invented in the last century (although Chinese linguistic textbooks going back many centuries explain syllables this way). This is the model for Bopomofo (used to write Taiwanese) and the Khom script invented by Ong Kommadam. (The ever-schizophrenic Wikipedia labels this model a "semi-syllabary" even though many of what it label "true" syllabic script like Japanese and Cree have separate graphemes for coda consonants and Sumerian has separate graphemes for rhymes.) Thus, syllabic script may be said to exist on a continuum between those that represent onsets and rhymes (e.g. Bopomofo), to bodies and rhymes (e.g. Sumerian), to bodies and rarely codas (e.g. Greek, Japanese, Mayan).
By comparison, alphasyllabic scripts have likewise developed multiple ways of representing speech. Alphasyllabic graphemes represent syllables or parts of syllables, but the individual phonemes are distinguishable in the same way as an alphabetic script whereas syllabic graphemes do not demonstrate any visual similarity between those with similar phonemic values. In the Brahmic scripts, consonants are written and then appended with diacritics representing vowels (a similar model is used by purely consonantal scripts as an aid for children and non-native speakers); most such scripts assume that an unmarked consonant has an inherent or implicit vowel (typically /a/) unless marked with a mute diacritic, but Thai (and Tolkien's Tengwar) indicates all vowels explicitly with diacritics and has no mute diacritic. In the Ethiopic scripts, the consonants are distorted in a less predictable fashion and thus more memorization is required. In the Meroitic script, the script was written like an otherwise alphabetic script except that all consonants were assumed to be followed by an implicit /a/ if not followed by another vowel letter; although in later stages of the language vowel letters had to be preceded by their associated glide if alone or starting a word. In the Canadian Aboriginal Syllabics script, which was invented in the last two centuries by missionaries, consonants are rotated to indicate a change in the vowel. (Old Persian cuneiform is sometimes classified as an alphasyllabic script, but it is actually a defective syllabic script.)
Most alphasyllabic scripts write independent consonants with dependent vowels and append the vowel to the preceding consonant (even if the syllable is a rhyme without onset), but there may easily be exceptions to this. Pahawh Hmong writes independent vowels with dependent consonants, while some modes of Tengwar append vowels to the following consonant (even without a coda).
The Paleohispanic scripts are a unique mixed alphabetic–syllabic script, as all plosives are represented by syllabograms since the language's phonotactics prohibit plosives from starting consonant clusters. The earliest iterations of the scripts were purely alphabetic but used multiple graphemes for single consonants depending on what vowel followed them, then discarded the redundant vowels once those graphemes were used to represent syllables. (Obviously, the ever-schizophrenic and outdated Wikipedia pages leave much to be desired.)
Alphabetic script and alphasyllabic script carry the same functional load (being sister systems and all) and are the most common type of script used by conlangers, presumably due to a mix of cultural bias (most conlangers speak languages written this way) and the smaller number of graphemes required compared to a syllabic script. Hangul is the logical extreme of this model, as all phonemes are written explicitly and neatly in syllable blocks that are more easily interpreted by the human brain.
For the largest archive of constructed scripts, I recommend omniglot dot com. I am currently maintaining a google sheet listing all the constructed or adapted scripts along with basic information like their type, direction and what language they are used to write to better sort and search them. The format is still a work in progress so I am not sharing it right now.