For example, take the Latin stare. Spanish changes it to estar. Status becomes estado. It looks like Spanish does not like the sound of ST in the beginning of the word. Or SC either; the Italian scuola is escuela in Spanish, studente is estudiante. Curiously, España is Spagna in Italian.
French does something similar (école and étudiant) even though they drop the S sound altogether in most of the words. "Spain" is Espagne in French.
I guess this is why some Spanish speakers have a hard time with "stop" (which some mispronounce as esstop) and this leads to overcorrection as in Ricky Ricardo's famous "Lucy, you got a lot 'splaining to do!"
Maybe it's Italian that is dropping the E sound ("to explain" in Italian is spiegare) but that does not explain the Latin stare or status.
Response 1: It has something to do with the sound systems we learn as infants. We discard those sounds we never hear as not relevant, and later can learn to produce them only with some difficulty. So the Latin-that-became-Spanish did not use combinations like st- as word initial. For the Spanish-speaker, a word-initial /st/ is about as challenging as some of those great clots of consonants one finds in a Polish or Czech word are to the English-speaker.
The real question is why the Latin-that-became-Spanish (and the Latin-that-became-French, apparently) lost the ability to make such sounds word-initial. Latin had many words like stare and schola, which become estar and escuela. I of course have done no research on this, and I don't think Penny takes up the question, but here's a wild speculation. As the Latin case-ending system breaks down, most forms of all words end in consonants. That means that it is increasingly rare to find the pattern -C C-; it's much more frequent to find -V C-. The exception is -Vs C-. Maybe all over Spain little children trying to systematize the language they are hearing conclude that where /s/ and /Consonant/ appear together, the combination is VsC. That is, their little sound-processing minds conclude that the sound in illa schola is /-as-/; that the sound in illas scholas is /-as-/; that the sound in illos stant is /-os-/....there are practically no environments in which they clearly hear /-sc-/ as word-initial. So they discard word-initial /s/+C from their inventories of relevant sounds.
This doesn't explain why the /-asC/ and /-osC/ they've heard merges to /-esC/, and that fact may show that my theory is all wet. But vowels are pretty fluid.
Response 2: Penny explains it this way (page 36): "Because the second consonant (typically p, t, c, qu or m) required maximum closure, speakers of Latin came to hear it [the second consonant] as syllable-initial, preceded by an anomalous 'semi-syllable' consisting of /s/. This difficulty would be particularly noticeable where the preceding word, if any, ended in a consonant (e.g. ad scholam) and was resolved by turning 'semi-syllabic' /s/ into a full syllable by means of an additional vowel (at first /i/, later to /e/)." As you point out, it happened in French and also in Portuguese, but it is curious that it did not happen in Italian. An unexplainable mystery, I guess.
Response 3: We tend to think of Italian as the ultimate Romance language because they were the first to be Latinized and Latin originated on the Italian peninsula, but there were notable lexical and grammatical influences by the Germanic tribes, who invaded Italy when the Roman Empire crumbled, on the languages of Northern Italy. But I do not have any conclusive examples on how this affects the main question of the thread.