Space filling versus optimal DOE

[#1]

For mixtures frequently space filling DOE's are used; guess these are not optimal? So what is then the benefit?

Why not use space filling also for non-mixtures instead of Optimal DOE? I am preparing a DOE course and expect these questions and so far I don't have the right answers. 

[#2]

Just to add to Victor's excellent explanations (although it may not be a specific answer to the OP), the methodologies need not be mutually exclusive.  It is possible to start with some knowledge of the model and investigate unknowns simultaneously and/or sequentially.  It is not unusual to start with hypothetical models based on subject matter knowledge, get data to support or reject those hypotheses and iterate to investigate alternatively (if hypotheses are not explaining/predicting the data well) or augmenting the space (if they are).  I suppose we are all looking for effective methods first and efficient methods second (e.g., optimality).  Much depends on what we know (or think we know) and then getting data to test that knowledge.  I think it is a good idea to "challenge" the model and determine where it fails, rather than just where it works.

[#3 · Accepted solution]

Completely agree with you @statman, depending on the objective and information about a product/process, the two design methods can be used sequentially : screen factors first with a model-based optimal approach, and then use a model-agnostic approach to optimize a predictive model in the eventuality of non-linear response.

Here is a slide from Synthace about the different designs and their possible complementarity :

There was also a paper on this topic of combining different DoE methodologies : https://community.jmp.com/t5/Discovery-Summit-Americas-2020/DOE-Gumbo-How-Hybrid-and-Augmenting-Desi...