Transmitting the experiment back at a loss (ex. transmitting Science Jr. experiments will only give 20%) will only give diminishing returns on the transmitted data (the 20%). So if you get 100 science from a Science Jr. on the Mun and then transmit only 20, then the next time you repeat it you will get 100-(20*0.71)=85.7 (0.71 is my calculated value for diminishing return). This means the next time you observe the Science Jr it will read 85.7 Science, and you can transmit back 20% of that or bring all 100% of it back home to Kerbin if you want to. So basically, if you have enough power (or solar panels) you don't lose anything by transmitting experiments, it just takes time to have to redo them so many times. When I went to Minimus with 4 Science Jr. and 4 Goo, I transmitted experiments, crew reports and EVA reports on both at high orbit, then at low orbit (gives different for each). This takes time since you have to do the experiments like 20 times each for the diminishing returns to reach 0 and waiting for solar panels to recharge. Then I landed and kept the samples from when I landed and returned to Kerbin. That mission netted me about 1200 Science, about 400 from the samples I brought back, and 800 from the transmitted ones!