Question re tone rules:

I understand the basics regarding the tone rules for the three consonant groups, but I'm a little confused about the following: When deciding the tone for a word without a tone marker, what is the determining factor; the vowel length or the final consonant? Take for example ถูก; Is it ถ = อักษรสูง + long vowel = คำเป็น or is it ถ = อักษรสูง + stop final consonant = คำตาย?

Question re clusters:

Am I right in saying that a cluster is "any two consonants written without a vowel"?

Without tonemarks, after the class of the initial consonant the tone is determined by how the syllable ends:

If a syllable ends with a vowel the vowel-length determines the tone, if it ends with a consonant the final consonant determines the tone.

So in your example ถูก it is high class initial consonant + stop final consonant = low tone (no need to worry about the vowel-length here)

The only exception is for syllables with low class initial consonants + stop final consonant. Here the tone is determined by the vowel-length: short vowel is high tone, long vowel is falling tone.

For instance วัด vs. วาด, วัด has a short vowel and is high tone, วาด has a long vowel and is falling tone

For the tone rules, it would be simpler for clearly enunciated speech if you assumed that a final short vowel implied a final glottal stop, which is an oral stop like /p/, /t/ and /k/. To that end, you can think of ะ as a final consonant (which is what it was originally - visarga, and still is in the Khmer script).

As to what a cluster is, it is very complicated and is actually a mess.

While technically a syllable-final consonant followed by another syllable (which necessarily starts with a consonant) is a cluster, it is rarely useful to think of it as one. It's occasionally useful to think of it as one when considering assimilation, such as colloquial ยังไง for อย่างไร.

A sequence of two consonants separated by a written vowel (when viewed in phonetic order) e.g. เล่น, or separated as initial and final, as in นม, is not a cluster. A sequence of two consonants at the start of a syllable, not separated by a vowel, is a cluster, as in ปลา [M]plaa 'fish', หมา [R]maa 'dog' or สระ [L]sa 'pool'. (สโรช below is actually a compound of the last word - the other element is ช, meaning 'born') A sequence of two or more consonants at the end of a word is a cluster, as in จักร​ [L]jak 'circle', เกียรติ [L]kiat 'honour' and เกียรติ์ [M]kian 'honour'.

The problem of terminology arises when the short /a/ with no vowel symbol occurs between the initial consonants, as in โสร่ง [L]sa[L]roong 'sarong', ถนน [L]tha[R]non 'road', เฉพาะ [L]cha[HS]phaw (or [L]cha[H]phaw) 'special', สระ [L]sa[L]ra 'vowel', นคร [H]na[M]khawn 'city', คณิตศาสตร์ [H]kha[H]nit[L]ta[L]saat 'mathematics', สโรช [L]sa[F]root 'lotus'. I would say the last three did not contain an initial cluster, but I have seen a claim that นคร contains an initial cluster. Some argue that there are two types of /a/ here - a normal short vowel /a/ and an epenthetic vowel as in the English pronunciation of 'Knorr' with /k/. I've yet to see a European linguist concede that the 'epenthetic vowel' in Thai is distinct from the ordinary short vowel. There's a lengthy discussion on clusters at Inside a Thai Syllable.