Determining whether a point is inside a circle is two subtractions, three multiplications, an addition, and a compare. ENIAC would not be overtaxed by that.
Exactly. All this comparing of hardware stats is totally beside the point where "grid versus non-grid" is concerned.
Say you've got a big fight with, I don't know, fifty combatants, auras, zones, and other "entities." Say your system requires you, every single turn, to calculate the distance between
every distinct pair of entities. This means you need to do the following:
1. Compute x1-x2, y1-y2, z1-z2 (I'm being extra demanding and requiring that our system can handle flying entities).
2. Take the square of each of those numbers.
3. Sum them and take the square root.
4. Do this 1,225 times*.
That's a total of 11,025 flops (floating-point operations, for those unfamiliar with the term). This is
nothing. By 1980, we already had chips capable of doing a megaflop per second. There are plenty of things you can do to a modern computer that will strain its hardware, but this ain't one of them.
As I said before, the potential benefits of using a grid would be to make things easier for the
developers, not the machines. And even there, the benefit isn't dramatic. A few tasks are easier, the UI is a bit simpler. But the great value of the grid to 4E was that it simplified old-fashioned, face-to-face battlemat play. The VTT would have followed that model, not because there was any compelling technical reason to do so, but to maintain compatibility with the tabletop game.
*50 entities; for each entity, there are 49 other entities it could be paired with; 50 x 49 = 2,450; divide by two because otherwise you're counting each pairing twice, once from each end.
