Mainly, he argues to install things (e.g. renewable capacity) only if they "economic and technical sense".

http://www.spiegel.de/international/germany/german-energy-expert-argues-again...

Kohler: The operator of a solar plant should be getting a message, namely that it's up to him to market the electricity he produces.

That sounds reasonable at first (and I agree with his general statement that we should only construct what actually makes sense). However, we really need a good and fair market system before you can really demand such thing from the system participants. The demand side gets guarantees, as well, right? I'm not entirely sure about the industrial demand, but I think it's pretty close to a connection guarantee if you open up a new factory or anything like that. And energy prices are not localised at all, so why should anyone bother thinking about location?
And why should it only be the supply side? In an optimal future setting, industrial demand also takes (local) energy and grid connection price into account when they decide where to locate. Then, the northern areas start to look better. Maybe cheap energy would even create job in Northern Germany and rural Bavaria.

He also is doesn't really fancy capacity market (same here), but instead says we should simply subsidise regular backup capacity:

Kohler: Anyone who guarantees the security of supply in the future has to be paid for it, even if his power plant is only needed at certain times.
SPIEGEL: Some receive subsidies for supplying green energy, while others are paid so that they'll be available in case it rains or the wind doesn't blow. It doesn't sound very market-based.
Kohler: It can indeed be organized in a market-based way throughout Europe by using so-called capacity markets. But that doesn't do any good. We have to synchronize the addition of more solar and wind energy systems with an expansion of the overall system, or the energy revolution will be a failure.

He closes with an interesting statement about the accuracy of energy forecasts:

Kohler: In the 1970s, they believed that there is an annual 6-percent linear increase in the demand for electricity. That number was used to estimate how many nuclear power plants had to be built. It was also the reason I went to work for the Institute for Applied Ecology (Öko Institut) in Freiburg at the time. I thought the calculations were fundamentally wrong. Today we have a solar and wind euphoria, instead of a nuclear euphoria. We believe that there will be a 10-percent decline in electricity consumption by 2020. And, once again, we assume that this change will be linear. But I'm sure that we're probably going to be wrong this time, too.