In some environments in Gk., PIE u > ü (maybe just dia.). I suggest that in all or most environments in Arm., PIE u > ü (since with further changes to neighboring C later, it would help explain some odd correspondences).
PIE ou > oy / uy
PIE eu > oy / uy
make the most sense if
ou eu
oü eü
oü öü rounding
öü öü fronting
öi öi simpl. 1
oi oi simpl. 2
This new front ü caused following K > KY in:
usanim = become accustomed to/learn, usuc’anem = teach Ar; uc- = be accustomed to/take pleasure in, okas = pleasure/dwelling-place S; vyknoNti = get accustomed to, uc^iti = teach OCS;
lu:x (n) lu:cis (g) = light L; loys lus- Ar;
yoke E; iugum L; zugón G; yugá-m S; luc Ar;
among others
and
if needed in:
doxtar P; dugëdar\duGðar- Av; dustr Ar;
This new front ü caused preceding T > TY in:
hardus Got; hard E; kratús = strong G; karcr = hard Ar;
cado: = fall L; cHacnum Ar;
among others
and
if needed in:
cHacnum either tn > tYnY or perhaps *cHacunum > cHacnum (since many in -num/nim etc. show the same V spreading, unetym. for PIE), though this is unclear.
Also, ü is seen in Nuristani and as a stage in Dardic (such as ü > we or ö > yo from umlaut). This presence in IA suggests that rúçant- = bright/shining S; could be caused by ü (perhaps reg. if only short u>ü and üK(>Y)V with analogy in most words).
Further, though spec., an Anat. change sim. to these could be seen in kYü > tu \ ut \ etc:
rúçant- = bright/shining S; {luta(i)- \ lutant-} = window Hit;
çrómata- S; hliumunt = fame OHG; hliuma = hearing / (pl) ears Got;
{tum(m)ant-} Luw; {is^tam(m)an[?]-} = ear, {is^tamas^-} = listen Hit;
with met in kYl- > ikYl- > ilkY- > etc. (other alt. of l / s is seen, though spec., in Anat. among others).
I’ll think further on which and how, if anything comes to mind on the uncertain ones.