Using data from VCP on sales since early 2006, I created two models for the two cards in question. I won't extrapolate the data, so we only have about 9 years to work with. I only used PSA 3,4, and 5 grades for simplicity, since the grade is such an important factor and looking at 4's vs 8's is a whole different ball game. Close grades should follow a closer regression fit. It's a pretty good set of data and a pretty good model fit in the end. I used a "curved" quadratic equation with dates and grades as my predictors (and the squared terms which end up having p-values of <0.001). R-sq for both cards are at 70%+. No signs the lead me to believe the model has errors (it is normally distributed, as expected)
Here's some stats on the 51 Mantle RC sales of PSA 3,4,5 via VCP.
Expected cost of PSA 4 in 2006: ~$2500
Expected cost of PSA 4 in 2015: ~$3900
9 year ROI: 56%
1 year ROI: 5.1%
Here's some stats on the '33 #53 Ruth sales of PSA 3,4,5 via VCP.
Expected cost in 2006: ~$3100
Expected cost in 2015: ~$4250
9 year ROI: 37%
1 year ROI: 3.6%
*I have a degree in mathematics and statistics, so I have a decent idea of what I'm talking about here...
While the Mantle is going up in price a bit faster, the Ruth has pretty good investment power as well and is in no way stagnant. To give value on something based on the future would be unwise.
*I watch Shark Tank every week and Kevin O'Leary reminds me every week not to invest in future forecasts and "what-ifs" 
Thanks for letting me "nerd-out" on my lunch break.