Basically, your music will only come out as good as your monitoring system will allow. It often amazes me how people are perfectly willing to drop thousands on modules but continue to monitor what they're doing with them on a $300 pair of powered crackerboxes. Or worse, headphones, which have proximity effect issues with low frequency sounds...mainly because they're not exactly suited for repro below about 300 Hz without some circuitry and/or physical design elements that will color up your results. Back many years ago, I was told to never, ever, EVER mix through headphones, and having experimented with that to see the reasons for myself, I know that that's something you shouldn't do.
But if you have a good set of monitors, set up correctly...then you'll know exactly what you're doing and can avoid all of these coloration issues you're noting. 98% of problems of this sort get fixed that way. Better still, if you have the ability to use multiple monitors, you can use one pair for uncolored, critical applications such as mixing, and then have a set of "crackerboxes" as well to use as "check" monitors, to see how the mix behaves in typical real-world situations. But you'd never want to actually MIX on the latter ones, instead just checking to see if something's glaringly wrong that needs correction.
Note that by "set up correctly", I'm including any acoustical treatments needed at your mixpoint or in your studio space in general. This especially goes for bass traps, which correct deceptive low-end buildup that results in a room with parallel walls that reinforce resonances known as "room nodes". If you don't know about these, go in a bathroom stall and start humming in the lower frequency range...and at some point, the stall will "ring" because you've hit a harmonic node of the space enclosed by the stall's walls. Just transfer what's going on with this experiment to your studio space, and you can easily see why treating studio spaces is just as critical as having a proper monitoring setup in it.
Secondly, this sort of problem is why program equalizers exist. These aren't the same as the more typical parametric or graphic EQs, but include such things as the Pultec EQP-1A which are designed for making broad coloration changes. They also tend to work differently, accentuating more than just the indicated frequencies on the controls. Along with a suitable compressor to merely ride gain, one of these belongs on your mixbus at all times, precisely to make large-scale timbral adjustments. I should also note that, when you're using ANY equalizer, the rule of thumb is to cut levels of objectionable sonic elements...not to boost everything to swamp them. And if the real deal here is too pricey (which it is!), try a good VST emulation such as Ignite Amps' PTEq-X...which is FREE (and on KVR Audio).
But again, without proper monitoring, you can have all the knobs for tweaks in the world and you'll STILL not have a good idea of what you're doing. Just like how you wouldn't drive around at night while wearing a smudged and scratched pair of sunglasses, you shouldn't be trying to mix on something not suited for the task. But 99 times out of 100, when people mention how all of their mixes are [INSERT PROBLEM HERE], the blame invariably comes back to the monitoring being used. Instead of looking for a synth-specific fix, or trying mixing techniques that're putatively for electronic sound (which, IMHO, don't work as advertised...I just mix electronic-based audio the same as any other large-scale multichannel mixdown), examine your monitors, how they're set up, what your workspace's room might be doing, and the like. In the long run, this will yield better, lasting, and consistent results.