Another problem with Spencer's theory is that he relies heavily on his own previous papers. That's never a good sign. You should be able to cite other scientists' work if there's a strong scientific basis to what you're doing.
One serious issue I have with his model is that he somehow allows for the positive PDO phases to have much more impact on global temperatures than the negative phases. Just compare his Figure 4 to the PDO index monthly values:
http://jisao.washington.edu/pdo/
How on Earth is he getting 0.6 deg C warming from 1900-1945, 0.3 deg C warming from 1975-Present, and just 0.2 deg C cooling from 1945-1975? That really seems like a stretch to me based on the PDO index values.
The other problem being that there was only a 0.45 deg C warming from 1910-1945 (even less if you go back to 1900), 0.1 deg C cooling from 1945-1975, and a 0.55 deg C warming from 1975-Present.
http://data.giss.nasa.gov/gistemp/graphs/Fig.A2.lrg.gif
This tells me that he's attributing too strong of a forcing to the PDO (actually cloudcover, indirectly impacted by PDO), because the amplitude of his modeled temperature changes are too large until the present, when he has to include the CO2 forcing to account for the recent warming.
I think if you look at Spencer's theory here with a truly skeptical eye, it's just not good. His model doesn't fit the data well and he's completely ignoring all but 2 forcings. Where is the solar forcing that deniers love so much? The problem is that Spencer is trying to attribute a greater forcing to PDO than it's actually responsible for, and to do so he has to neglect other forcings (like the 1 W/m^2 solar forcing from 1910-1940). Just like Scafetta&West ignore all other forcings in order to attribute a greater effect to the Sun than it's actually responsible for.
And they get similar results. Even with their unrealistic models, Spencer can only attribute about one-third of the recent warming to PDO, and Scafetta&West can only attribute about one-third to solar effects. And their models directly contradict eachother by ignoring the other's forcing.
Spencer's PDO+CO2 model isn't great, but it isn't bad. But if he made a PDO+CO2+solar model, it would be utter garbage.
Yet another issue I just noticed is that Spencer's PDO-only model has a greater warming than his PDO+CO2 model in the eartly 20th century. What the hell is that?? Increasing CO2 causing cooling?? That makes no physical sense whatsoever.
Deniers dismiss any work done with climate models as totally speculative and worthless - unless it's done by a 'skeptic' like Roy Spencer or Scafetta&West. It doesn't matter that it's an (admittedly) grossly oversimplified model. As long as it supports the denial point of view, it's gold.
But to answer your question, the reason 'skeptics' love the PDO is the same reason they love the Sun and GCRs. It's not anthropogenic!