Am able to influence the horizontal linearity using magnets.
Put some on L3 . Still not perfect but can try other magnets or location.
Reversing the polarity on L3 cannot have any effect if C1 is truly healthy, since the job of C1 is to block dc. If there's no dc, then L3 behaves the same regardless of which way it's inserted. Since you replaced C1 with a high-voltage cap, I'm inclined to absolve the capacitor of any role in producing nonlinearity, but that also makes me doubt that anything changed at all when you reversed the connections on L3.
It's hard to tell from a nearly blank display field whether anything changes, and by how much. You might want to put up a page of text filled with characters. I wonder how much of what you think you see is actually there. You will want to know if the linearity gradually shifts in going from left to right, or if there is a more abrupt change (i.e., linearity is good within both halves, but the two halves don't match). Details like that are useful clues as to what could be wrong.
Putting magnets on L3 will certainly have an effect, as you've noticed, but that's not a cure. You need to find what's causing the nonlinearity in the first place. And the first step before that is to characterize it more carefully so that you can quantify any fixes that you implement.
Standard things to try first include touching up all the solder joints on J1, the flyback transformer, and components associated with those. Check CR1 and CR5 and their connections. Resistances in all of those connections can cause nonlinearity. There's also a cap from flyback pin 4 to +12V; check it, too [pin numbers are from my Compact Mac Repair doc; I don't think there are any numbers silkscreened on the PC board]. In other words, check those obvious things first before applying band-aids.