But I agree that overly large phonemic inventory is a problem. On the other hand, it seems that languages either have a complex consonant system but a simple vowel system; or a simple consonant system but a complex vowel system; I haven't yet seen a language where both systems are simple (Japanese vowels have tonality, so it's not a simple system IMHO), probably because the words in such a language would have to be quite long.
There are quite a few languages where both vowel and consonant systems are simple - just look at Polynesian languages such as Māori. The latter's vowel system is 5-vowel, and "long vowels" are phonemically vowel sequences that span moras. But, yes, it does mean that you end up with long words such as "whakararurarutia". That said, it's a rather extreme case, and one can still construct fairly simple but rich consonant systems in practice, because it's basically combinatorics - adding just one more bit of information doubles the domain space! So e.g. if you start with a strict CV consonant system and allow C(l/r)V, that's almost 4x as many contrasting syllables. Make it C(l/r/w/y)V(C) like in Globasa, and even with considerable restrictions on clustering stops etc this is enough for most words to be 3 syllables or less, and for most function words to be 1 syllable.
Now, the real success of Esperanto is that it does have an over 1 century international active community that does produce it’s own cultural artifacts, using Esperanto as a communication mean. All that without a bound army to back it at any point, that’s probably an unique feat in human history. Also to make it clear, it was not meant to be a universal language, but an international one.
Personally, I love that projects like Globasa comes to live. On a pragmatic level, large scale adoption is unlikely, but that is the case of any human endeavor. Let’s make sure that grandiloquence result likeliness never inhibit beautiful dreams being pursued.
That said, it's still a massive improvement on English phonologically. Even if you only consider the simpler American varieties, the three-way æ/ɐ/ɑ distinction alone (as in bat vs but vs bar) is a huge WTF for anyone coming from a typical 5-vowel system. And then you have consonants like θ and ð that don't have clear 1:1 counterparts in most other languages, often not even as allophones of something else that you could point at.
Still, if you want to see what a more modern take on the concept might look like, I believe Globasa (https://www.globasa.net/eng) is the most active project along those lines. Of course, realistically, the likelihood of it actually being adopted as the universal language is effectively nil, but then that's also the case for Esperanto.