Have there ever been (serious) attempts to replace alphabets with phonetic alphabets?

I think you're confusing the issue a little bit. You need to differentiate between the concept of phonetic alphabet and transparent orthography.

The introduction of the IPA or international phonetic alphabet was the one more or less successful attempt at making communication among linguists and language educator more consistent across languages. But most of its uses focus on key phonemic (rather than phonetic) features of the individual language and are subject to convention. So the pronunciations in a dictionary will not reflect many phonetic features like the nazalisation of vowels, etc. These conventions often compete with language change and dialectal variation. Take for instance the 'a' in English 'bat' Its pronunciation varies greatly across English dialects but in Britain has mostly moved from /æ/ to /a/. However, in most neutral contexts /æ/ is still used for consistency's sake. A true phonetic transcription would also record assimilations leading to spelling such as 'books' and 'budz'. To make things even more complicated, many sounds that are recorded by the same IPA symbol across languages do not actually sound exactly the same - this is particularly true for vowels - leading to more confusion and the necessity to learn local conventions for each language when reading transcriptions for the purposes of cross-linguistic comparison.

As a result an adoption of a fully phonetic alphabet that would apply to all languages or even just one language was not ever contemplated because there would be no advantage to it.

However, transparent orthography is a much more sensible concept. All it indicates that the way words are written makes it easy for native speakers who learned to associate individual sounds with letters (or groups of letters) to reliably pronounce any string as it was intended. No language is completely transparent in the sense, that it is necessary for readers to learn some conventions about how to pronounce certain letters in context, but there are languages that are almost completely transparent (Italian, Slovene) and those that are almost opaque (Mandarin) or very opaque (English) or somewhat opaque (French). Some othographies are only opaque in one direction (Greek - easy to read, more difficult to write, Arabic - easier to write, more difficult to read).

Many languages have successfully reformed their orthographies to be more transparent but English (which is what I think prompted your question) has not. There were some attempts to do this in the US in the 19th century which gave us things like color/colour but they only tinker around the edges. There are several reform movements active in English orthography but they have no prospect of being successful in the short-to-medium term.

Any possible advantages of transparent orhographies are purely marginal. Opaque orthographies have not prevented cultural, economic or geopolitical success of cultures for which they are typical (see US and China). They have certain obvious disadvantages for learning and use but the transactional cost is so small overall, that the possible long-term gain does not outweigh the huge short-term cost of a switch for a large and powerful culture like the Anglophone world or China. Smaller cultures have often undergone such reform but it was always linked to larger issues of cultural and political identity - never purely because of some sort of rationally judged advantage of a more transparent system.