This is a very good question, and I've been unable to answer it. From what I understand, suppression happens because of overstimulation or understimulation of receptors.
For example, testicular atrophy happens when there's a lack of LH and FSH --- the LH receptors don't get any stimulation, and the leydig cells die. This is an example of understimulation. On the other hand, if you inject very large, frequent doses of HCG, the LH receptors on your testes can get overstimulated. Thus, when you try to go natural, the testes are insensitive to the normal levels of LH produced by the pituitary, and low T results. This is an example of suppression from overstimulation.
The question becomes: will the hypothalamus experience bad effects from having its androgen receptors (or ERs for that matter) overstimulated? If these receptors became desensitized, would not the hypothalamus become insensitive to T, and thus "think" that T was low? From what I understand, high levels of T result in AR upregulation, so hypothalamic desensitization from elevated T doesn't seem likely. However, could there be other deleterious effects from the hypothalamus having not secreted GnRH for long periods of time? Given that this lack of GnRH secretion is not a result of the hypothalamus being understimulated, I don't have a good understanding of whether or not the hypothalamus will have a hard time coming back online.
One thing seems true though: if the hypothalamus would have a hard time coming back online after a steroid cycle involving GnRH, it would also have a hard time coming back online after an ordinary steroid cycle. I say this because from the hypothalamus' perspective, the situations look the same: too much T and E2. I don't think the hypothalamus would change its behavior based on the presence of exogenous GnRH, because I don't think the hypothalamus can directly detect the GnRH. Thus, we should be seeing some cases of steroid-induced hypothalamic hypogonadism, despite the dearth of athletes using Kryptocur. The fact that we're not seeing these cases makes me think that the hypothalamus doesn't typically suffer from steroid use. However, I can't say this with any certainty...
Yes, I agree that natural release of GnRH would be suppressed, but from my understanding, the rest of the HPG axis would be stimulated from the exogenous GnRH, which seems like it would prevent secondary and primary hypogonadism. Without the GnRH stimulation from Kryptocur, secondary hypogonadism remains a possibility, even with HCG use...
Scally brought up the feedback path to the pituitary. Exactly what the pituitary will do given these conflicting signals (GnRH tells the pituitary to increase gonadotropins, while elevated T/E2 tells the pituitary to decrease gonadotropins), I cannot say (and he chose not to illucidate). My assumption was that the pituitary would secrete gonadotropins if either one of these signals was positive. However, this is speculation on my part. Any ideas?