For the sake of fun lets look at Devils Canyon. It appears to be capable of an output of 18,000 GW-h per year. According to the EPA in relation to the above graph:
"an increase of 1 unit on this graph (1 x 1022 joules) is equal to approximately 18 times the total amount of energy used by all the people on Earth in a year.
4"
And according to the IEA global power consumption was 157,485 TW-h in 2013. So now we have a few numbers that we can work with here we can say that it would take 8749.16 Devils Canyons to supply all forms of energy for humanity for the year of 2013.
So in order to cause the oceans to tick up a single 1e+22 joule of energy would require 157,485 Devil's Canyons with a 100% efficient heat pump going into the ocean. In addition to that, the assumption would be that the ocean also isn't radiating out that additional energy as well so the amount of energy required would be far far greater.
Onto the next point, you shouldn't see too much correlation as the paper specifically stated that:
"The maps above show how the reflectivity of Earth—the amount of sunlight reflected back into space—changed between March 1, 2000, and December 31, 2011. This global picture of reflectivity (also called albedo) appears to be a muddle, with different areas reflecting more or less sunlight over the 12-year record. Shades of blue mark areas that reflected more sunlight over time (increasing albedo), and orange areas denote less reflection (lower albedo).
Taken across the planet, no significant global trend appears. As noted in the anomaly plot below, global albedo rose and fell in different years, but did not necessarily head in either direction for long."
So basically it appears to be a wash in terms of what the global albedo looks like in the past nearly 12 years although they do note that additional information is required.
Regardless if the issue was with the heat island effect or cirrus clouds it would have a much more visible impact. Basically neither of your proposed explanations hold up to measurements.
