This extends to other words that must agree. Instead of thinking "the noun is masculine so the adjective must be masculine as well", think "the article is 'il' therefore the adjective is 'buono' instead of 'buona'".
For non-Russian speakers, the two letters get the sound that you expect if stressed, and otherwise sound like а. This rule also applies to borrow words that were transliterated into Cyrillic. So the English computer becomes компью́тер, and the stress goes on ...пью́т... (the English ...put... bit of the word). As a result that first о became an а sound.
My wife's reassurances that Cyrillic is phonetic likewise didn't work out for me. You can't pronounce the written word correctly without knowing where the stress is. You can't write down the spoken word correctly without knowing which unstressed а sounds are written as о.
Of course this is far better than English spelling...
I always felt like Russian was a pretty easy language to learn because it was so regular. Yes there are a lot of cases and declensions, but once you learn the rules, you can get like 95% of the way there, and then even the last 5% of exceptions are quite "regular exceptions", e.g. the "ogo" written -> "ova" pronounced rule.
It didn't work for me in Mandarin though, where I can recall the sounds of words but not the tone.
No one is out there watching to make fun of you if you count horses in 只 instead of 匹.
I used to have this mindset about german, oh who cares about grammar. it really limited me at upper intermediate level.
Chinese measure words are not something that happens on a noun-by-noun basis. Memorizing them as part of a noun is not a well-chosen approach. Most often this will saddle you with the burden of separately memorizing "pass" and "passed" as independent forms of the verb, while providing comparable benefits.
You might have noticed that my comment above suggested that
(a) the default measure word is 个;
(b) the default measure word for a horse is not 个.
Do you know why?