Okay, I went and did some math. Maybe this will help clarify my viewpoint.
In 2009 - 2010, the NHL salary cap was $56.8m. A second pairing defenseman (let's say this is the most likely outcome for Mattias - a solid second pairing shut down guy in a vaccuum that plays on the top line w/Dahlin) that year made around $3m. That $3m represented 5.3% of the total cap. Currently, Samuelsson's contract at $4.29m represents 5.1% of the total cap next year at $83.5m.
In 2029, if the cap goes up roughly $2m per year, we will have a cap of $96m. I think both of those are fairly reasonable numbers to project. If we split the difference between the two contracts at 5.2% - that projects out to the average second pair defenseman making $4.992,000. Representing a savings of $702,000 for Samuelsson's contract - in 2029. While it's certainly not nothing, calling this contract a "steal" in any sense isn't really accurate. It also ignores the inherent risk (however likely one deems that to be) that comes with offering that contract before he's proven anything.
Could Samuelsson end up being more than a solid shut down guy - and include some more offense? Sure - but I don't think that's what anybody is necessarily expecting. It seems like we've set the expectation at around 20-30 points - which I think is perfectly in-line. If he consistently puts up more than 30 points - yes, he's worth more than that. If not, we're probably paying him a bit too much for the next handful of years, and getting some (small) savings in the latter years - assuming everything goes according to plan. If it doesn't - then it's a horrible contract. There inlies the risk/reward people wanted to talk about. We take all of the risk, for a fairly small reward - barring a somewhat unforseen offensive game that Samuelsson would provide.