Add a norm parameter to the EBL model

[JouvinLea]

This PR adds a norm to the EBL model to give the possibility to the user to add a norm for the AbsorbedSpectralModel class and also to let it free in the fits.

It is also adding a small argument for the peek() method of FLuxPointDataset because for now the uncertainties are not handle properly and it crashed if you try to plot the error of this model.

[cdeil]

@JouvinLea - Thanks!

Currently it's not really clear how the AbsorbedSpectralModel uses Absorption:
https://docs.gammapy.org/0.14/api/gammapy.modeling.models.AbsorbedSpectralModel.html
Now that there is an alpha parameter, it's even less clear.
@JouvinLea - Maybe you could add the formula that is applied, and that would make it crystal-clear already, without needing much description?
Is there a good literature reference with this formula for absorption?
Even if it's simple and standard, I think for people like me that have never looked at EBL stuff it would be useful.

I'm not a fan of the plot_error option you're adding. IMO this not working should be considered a bug in Gammapy and we should just fix it (#2255). Would prefer not to add such workarounds in Gammapy. But OK, if you want it, because those plotting functions are important to you for this model, then we can merge this in, given that it's not clear if / when / who will fix this.

[JouvinLea]

@cdeil
I fixed the small docs and space issues.
I added a description for the Adsorbed model, hope it is clearer!

Personally I would prefer to stay with this plot_error parameter. I put a TODO to say to remove it when the issue is fixed. I know exactly which line to comment to have my script running but this is a useful plot routine so I think for non developper it is better to have this option. But as you wish, I can comments those lines by myself each time I run my script.

[cdeil]

@JouvinLea - Please fix:
https://dev.azure.com/gammapy/gammapy/_build/results?buildId=2035&view=logs&jobId=d7d66435-5ae4-561a-5876-2cd4e19b3ce2&taskId=922dfe32-5b90-53bc-18db-0fa0ecbbd97b&lineStart=187&lineEnd=188&colStart=1&colEnd=1

[cdeil]

@JouvinLea - Thanks!