The models presented for XP buy-in so far look like disasters waiting to make land-fall. With those models eventually the math plays out so that it becomes empirically superior to dump 10,000 XP into a handful of multi-class levels instead of a fraction of a single higher level - even if it is just for the sake of hit dice and attack bonuses.
No, if multi-classing even vaguely resembles what the articles have already described then it is going to be a hard level-for-level trade. The only way to make it work is to have the worth of multi-classing escalate based on two variables: 1.) Total number of levels in [Multi-Class Class XYZ] and 2.) Total character level - IE F10/(MC)MU2 = Level 12.
Basically having 1 level of Multi-Class Magic-User means something when you are F1/(MC)MU1 (Level 2) and something different 4 levels later when you make F5/(MC)MU1 (Level 6). F4/(MC)MU2 should be demonstrably different from F5/(MC)MU2 as well, but completely comparable in power level due to both being 6-level builds.
I can't see any reliable scaling math that holds up over time free-range. I think they'll at least need to be a constraint that you can't have more Multi-Class levels than your primary class levels. That might at least make things manageable. Even then, though, to make sure things scale relevantly with level trades you'd need multiple tables: one for each level of Multi-Class (Whatever Class), with an relationship of character level vs. benefits.
For example:
<table> <caption>Multi-Class Cleric Level 1</caption> <thead> <tr> <th>Character Level</th> <th>Benefit</th> </tr> </thead> <tbody> <tr> <td>2</td> <td>Channel Divinity 1d8/day</td> </tr> <tr> <td>3</td> <td>Channel Divinity 1d8/day, 1 Minor Spell</td> </tr> <tr> <td>4</td> <td>Channel Divinity 1d8/day, 1 Minor Spell, 1 First Level Spell /day</td> </tr> <tr> <td>5</td> <td>Channel Divinity 1d8/day, 1 Minor Spell, 2 First Level Spell /day</td> </tr> <tr> <td>6</td> <td>Channel Divinity 2d8/day, 2 Minor Spells, 2 First Level Spells /day</td> </tr> </tbody> </table>
Etc. Ad Nauseam up to Character Level 19.
Then you'd need another table like this:
<table> <caption>Multi-Class Cleric Level 2</caption> <thead> <tr> <th>Character Level</th> <th>Benefit</th> </tr> </thead> <tbody> <tr> <td>4</td> <td>Channel Divinity 2d8/day, 2 Minor Spells, 2 First Level Spells /day</td> </tr> <tr> <td>5</td> <td>Channel Divinity 2d8/day; 2 Minor Spells; 3 First Level, 1 Second Level Spells /day</td> </tr> <tr> <td>6</td> <td>Channel Divinity 2d8/day; 2 Minor Spells; 3 First Level, 2 Second Level Spells /day</td> </tr> <tr> <td>7</td> <td>Channel Divinity 2d8/day; 2 Minor Spells; 3 First Level, 3 Second Level Spells /day</td> </tr> <tr> <td>8</td> <td>Channel Divinity 2d8/day; 2 Minor Spells; 3 First Level, 3 Second Level, 1 Third Level Spells /day</td> </tr> </tbody></table>
All the way up to Level 18. Then you need a Multi-Class Cleric Level 3 table that runs from levels 6 - 17, a Multi-Class Cleric Level 4 table from 8 - 16 ... up to a Multi-Class Cleric Leve 10 entry that's only one row: Character level 20: Benefits comparable to losing levels 11-20 of your primary class.
Like I said, metrics that don't create obvious scenarios where Multi-Classing is categorically better or worse than playing a class straight are going to be ridiculously tightly controlled data-point by data-point, not some willy-nilly combination scale like 1st Ed or 3rd Ed used.
1E scale was weak sauce except for level caps: two CL3/MU3 in a party were nowhere near as powerful as one CL6 and one MU6 (yes, I know I have the XP by class scale over-simplified).
3E scale was broken except that dedicated casters were Quadratic and thusly head-and-shoulders better than any non-casting multi-classer while multi-classing was terrible for casters.
Let's face it, Dungeons and Dragons has never had a balanced, elegant, or effective multi-classing system. The most popular variants were only held together by existing rules defects that nobody wants to see come back. Doing it right is going to be ridiculously difficult and detail-oriented work.
- Marty Lund