It is a real pleasure to discuss these things with someone like you who has a great skill in language/grammar..  

yıkanıldı has exactly the same meaning as peseydyttiin does in Finnish: ´some people (probably more than one) washed themselves in the past.´

To the contrary of what some people may think, reflexive and passive are different things and a reflexive verb can have its passive form when necessary.

Remember the example I gave in my previous post:

a) Böyle de giyinmez ki (reflexive)

b) Böyle de giyinilmez ki (reflexive+passive)

This example explains well that reflexive+passive in Turkish  has its field of use. 

We also have verbs containing reflexive+passive combination:

aranıl(mak), yüklenil(mek),etc. 

..and an adjective: 

kaçınılmaz

Maybe there are a few more but these are what I remember at the moment.

The pleasure is all mine.

Reflexive + passive certainly has its function in language as your examples show, scalpel. But still it doesn´t erase the fact that there is always a chance for analogy also. We can say it is even probable that these forms spread wider than they were actually ment for by means of analogy, just like gokuyum said earlier. Because for an average language user it is a choice between models: real models and invented models, understood and misunderstood models...

Wow, this turned out a nice thread eventually (even though at some point I thought I messed it).

"There is always a chance for analogy also" .. Yes, there is and we can´t control it.. As you said "an average language user"  may be tempted to form new words by analogy.. I have a dictionary published in 60´s  and it says ´ya da´ is a badly made-up thing in place of ´veya´ and shouldn´t be used, and now in 2011 almost we all use ya da and veya is already an archaic word..  

Thank you Abla for supporting my idea. I think analogy is the main reason of this situation. Some people subconsciously think there is only one way to do passive and they use "n" as a buffer letter. I also want to thank scalpel too. Once upon a time he recommended me Feyza Hepçilingirler´s Turkish grammar book and I liked the book a lot and started to understand Turkish grammar better and think about it. 

The primary function of -n- is forming reflexive verbs often from transitives and reflexives are often intransitive: al-ı-n, sal-ı-n, gez-i-n, giy-i-n, dola-n, tıka-n, bul-u-n, aç-ı-n, öğre-n, tut-u-n, döv-ü-n, sakla-n, ara-n, taşı-n, sür-ü-n, söyle-n, kaç-ı-n, mırılda-n, tap-ı-n, etc. It is one of the most widely used suffixes from verb to verb.

The secondary function of -n- is forming passive verbs where the suffix -l- (which is the genuine suffix for passive) is not possible (verbs ending in a vowel or consonant -l)

ara-n (arandı belasını buldu), söyle-n (söylene söylene gitti), süsle-n (kadın süslendi), sil-i-n (silinip kurulandı ) are reflexive.

ara-n (ev arandı ), söyle-n (söz söylendi), süsle-n (şehir süslendi), sil-i-n (camlar silindi) are passive.

Either because of (as you said above) -n- is not considered as a passive suffix, or because to avoid confusion between the two uses, we can add -l- to -n-: başla-n-ı-l, de-n-i-l, ye-n-i-l, bul-u-n-u-l, oku-n-u-l,ara-n-ıl, söyle-n-i-l,etc.

But what made this thread turn into an interesting one is something else.. We shared our opinions about the "reflexive+passive" combination and its possible field of use..      

Average language users have surprisingly strong opinions about what is right and what is wrong and they are motivated to make right choices. The problem is that few people can really analyse the flow of language. I can imagine that the double role of -n- in derived Turkish verbs may be a reason for speakers to make further differentiations.

A good discussion is a rare thing. We people tend to take things personally and stick to our old views stubbornly. It concerns each one of us. When someone has typed half a page to make you understand a new thing the first reaction  -  to be honest  - is "what is he trying to prove, I know this better, who does he think he is". It is a protection mechanism which we need to get along but still a great obstacle to any learning.

Hmm. OK.

We can also make reflexive construction as in English.

Kendisini yıkadı = He washed himself.

In that case passive can not be used.

I wonder why. kendi can´t take the place of the subject or what?

Kendisi yıkandı.

Here "yıkandı" should be passive but concieved as reflexive.

Kendisi yıkandı = He washed himself by himself.

Aaa...like we got to the same point again.

I have visited the wikipedia page again:

In most languages that allow impersonal passives, only unergative verbs may undergo impersonal passivization.  Unaccusative verbs may not.  The ability to undergo this transformation is a frequently used test to distinguish unergative and unaccusative verbs.  In Turkish, for example, the verb çalışmak "to work" is unergative and may therefore be passivized:

Burada çalış-ıl-ır.

here work-PASS-PRESENT

"Here it is worked."

The verb ölmek "to die", however, is unaccusative and may not be passivized:

*Burada öl-ün-ür.

here die-PASS-PRESENT

"Here it is died."

However its conclusion about unacusative verbs is not the case in Turkish.

For example "burada ölünür" is meaningful in Turkish.

Burada ölünür = "here it is died" (literally not meaningful in English)

meaning this is the place one can die (directly) or

this is the place one should live until he dies (indirectly)

Similarly you can say:

Burada doğulur = This is the place to be born