I would like to continue on [question I][1] asking, what is a homotopy between 
Lie infinity algebras, since I'm not satisfied in two directions:

1.) The naive approach to define a homotopy would be ('naive' in my opinion 
of course) the following:

Let $(L,(l)_{k\in\mathbb{N}})$ and $(M,(d)_{k\in\mathbb{N}})$ be two Lie infinity
algebras, let $f_\infty,g_\infty:L \to M$ be two morphism 
(in the most general sense) and let $(C(L),Q_L)$ and $(C(M),Q_M)$ be the 
appropriate differential graded coalgebras with induced morphism 
$F,G:C(L)\to C(M)$. 

Then a **homotopy** between $F$ and $G$ is a degree $+1$ map

$H:C(L)\to C(M)$ such that $F-G = HQ_L \pm Q_MH$

let signs and additional structure of $H$ (linear, coalgebra, ...) aside for
a moment.

It is strange, however, that I never saw this approach in the literature. Is this definition of homotopy equivalent to the previously mentioned approaches
in [question I][2]? 

Now for what is more important:

2.) The homotopy theory of Lie infinity algebras as given by Urs Schreiber in 
[question I][3] is obtained by 'transferring' the homotopy theory of differential graded Lie algebras 'along' the following adjunction:

We have the functor $R$ from Lie infinity algebras to DG Lie algebras, that 
projects the Lie infinity algebra onto the homology of its underlying chain complex and then forgets the higher brackets and on the other side we have the 
functor that includes a DG Lie algebra into Lie infinity algebras 
(because every DG Lie a is in particular a Lie infinity algebra). This gives 
the adjunction. Am I right here?

BUT, in general $R$ loses a lot of homotopy information! This was for example made clear in the work of [Loday & Vallette][4] on operads. In particular, the complete homotopical information is transferred by the **homotopy transfer theorem** and $R$ is just a 'low degree shadow' of this, so to say.

Now the question that really irritates me for quite some time is: how can we be sure that we get the correct homotopy theory of Lie infinity algebras by transferring its 'shadow' in the category of DG Lie algebra back along the previous mentioned adjunction? To me it looks like we can not rule out that there is a more general definition of weak equivalences in the category of Lie infinity algebras, which just project under $R$ onto those we already know. 

Sorry if the second question is vague. 

If I had to set-up a homotopy theory I would (again naively) say, that 

- The homology is not (only) that of the underlying (co)chain complex 
$(L,l_1)$ but the
homology of the (co)chain complex $(C(L),Q_L)$ instead. 

- Quasi-Isomorphisms are those coalgebra-maps that become isomorphisms on this extended homology of the coalgebra.


Edit; I used the 'Lie algebra cohomology tag' since the homotopy theory of
Lie infinity algebras affects Lie algebra cohomology, too. 


  [1]: http://mathoverflow.net/questions/139175/what-is-a-homotopy-between-l-infty-algebra-morphisms "Question I"
  [2]: http://mathoverflow.net/questions/139175/what-is-a-homotopy-between-l-infty-algebra-morphisms "Question I"
  [3]: http://mathoverflow.net/questions/139175/what-is-a-homotopy-between-l-infty-algebra-morphisms "Question I"
  [4]: http://math.unice.fr/~brunov/Operads.pdf