About pronouncing the 's' in plural nouns

Solution 1:

Yes to both your questions.

Native speakers really do follow that “rule” as to whether being next to a vowel or a voiced consonant makes the -s suffix voiced as well, whereas being next to an unvoiced consonant makes the suffix also unvoiced.

And secondly, the same phonologic law is in operation when constructing a third-person singular verb.

You can also add a third class to that: forming possessives with apostrophe-s.

All three work the same soundwise.

Solution 2:

As far as I know, there are no notable groups of speakers that don't follow this rule phonologically: pens and pence, tens and tense have different pronunciations for all speakers with typical accents.

One thing that might be causing confusion is that the phoneme /z/ in English is not always fully phonetically voiced. A word like tens might be pronounced with a voiceless or partially voiceless fricative, but it still would sound distinct from tense because the vowel in ten and tens would be phonetically longer than the vowel in tense. For more details on this, see Are "whores" and "horse" homophones? and The pronunciation of ending "s".

For a fair number of English speakers, there is a merger in perception (possibly also in production, although I don't know if studies have shown whether it is a complete or partial merger) between /ns, ms/ and /nts, mps/ in many contexts, including word-finally. A smaller number of speakers may merge /ls/ with /lts/. I suppose that, if we analyze the merged pronunciation as containing the phoneme /t/, it would be formally possible to analyze the tens/tense contrast as a phonemic contrast of /ns/ vs. /nts/, but I don't think that would be a very convincing analysis.

I haven't heard of any English speakers merging /rs/ with /rts/.