actually the flaw in the radio that lets LO signal bleed out is that the mixer has poor isolation between the LO input port and the IF input port. to keep the LO out of the antenna, the mixer stage needs to be isolated better. this costs a few pennies more than manufacturers are willing to part with, so they go with the cheaper mixer stage that provides just enough isolation that the radio still works. the best isolation would be from the use of a Doubly Balanced Mixer. the LO isolation in a DBM can be as good as -70 or -80 db. the cheap mixers using emitter and base of a transistor as the input ports only get -20 to -30db isolation. the simplest DBM is 3 balanced rf transformers, and a diode ring. depending on how well you match the diodes (Vf and Cj) the isolation between input ports can be as good as -60 db. a DBM using transistors can get even better results, especially with matched transistors. ironically, it was spurious RF emission from radio receivers becoming a big problem that was one of the reasons that superhet radios became the standard, regenerative receivers radiated detector signals with almost no isolation at all, and this was a BIG rfi problem. some countries even passed laws requiring all radio receivers sold to the public be of a superhet design. this didn't do away with the problem entirely, but did reduce it a lot. so now there are cheapie superhet devices that have -30db isolation (which is better than no isolation) but there are ways of improving on this, but the only people that will improve on it are engineers that see a need for it. if i were to build my own receiver, you betcha i would use a mixer with high isolation. not only does it reduce the LO emissions, but it improves the performance of the receiver drastically.